Managing File Systems in Oracle Solaris

11.4

Part No: E61016

November 2020
Managing File Systems in Oracle Solaris 11.4
Part No: E61016
Copyright © 2004, 2020, Oracle and/or its affiliates.
License Restrictions Warranty/Consequential Damages Disclaimer
This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except
as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform,
publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is
prohibited.
Warranty Disclaimer
The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing.
Restricted Rights Notice
If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable:
U.S. GOVERNMENT END USERS: Oracle programs (including any operating system, integrated software, any programs embedded, installed or activated on delivered hardware,
and modifications of such programs) and Oracle computer documentation or other Oracle data delivered to or accessed by U.S. Government end users are "commercial computer
software" or "commercial computer software documentation" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental regulations. As such, the
use, reproduction, duplication, release, display, disclosure, modification, preparation of derivative works, and/or adaptation of i) Oracle programs (including any operating system,
integrated software, any programs embedded, installed or activated on delivered hardware, and modifications of such programs), ii) Oracle computer documentation and/or iii) other
Oracle data, is subject to the rights and limitations specified in the license contained in the applicable contract. The terms governing the U.S. Government's use of Oracle cloud
services are defined by the applicable contract for such services. No other rights are granted to the U.S. Government.
Hazardous Applications Notice
This software or hardware is developed for general use in a variety of information management applications. It is not developed or intended for use in any inherently dangerous
applications, including applications that may create a risk of personal injury. If you use this software or hardware in dangerous applications, then you shall be responsible to take all
appropriate fail-safe, backup, redundancy, and other measures to ensure its safe use. Oracle Corporation and its affiliates disclaim any liability for any damages caused by use of this
software or hardware in dangerous applications.
Trademark Notice
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Intel and Intel Inside are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and are trademarks or registered trademarks of
SPARC International, Inc. AMD, Epyc, and the AMD logo are trademarks or registered trademarks of Advanced Micro Devices. UNIX is a registered trademark of The Open Group.
Third-Party Content, Products, and Services Disclaimer
This software or hardware and documentation may provide access to or information about content, products, and services from third parties. Oracle Corporation and its affiliates are
not responsible for and expressly disclaim all warranties of any kind with respect to third-party content, products, and services unless otherwise set forth in an applicable agreement
between you and Oracle. Oracle Corporation and its affiliates will not be responsible for any loss, costs, or damages incurred due to your access to or use of third-party content,
products, or services, except as set forth in an applicable agreement between you and Oracle.
Pre-General Availability Draft Label and Publication Date
Pre-General Availability: 2020-01-15
Pre-General Availability Draft Documentation Notice
If this document is in public or private pre-General Availability status:
This documentation is in pre-General Availability status and is intended for demonstration and preliminary use only. It may not be specific to the hardware on which you are using
the software. Oracle Corporation and its affiliates are not responsible for and expressly disclaim all warranties of any kind with respect to this documentation and will not be
responsible for any loss, costs, or damages incurred due to the use of this documentation.
Oracle Confidential Label
ORACLE CONFIDENTIAL. For authorized use only. Do not distribute to third parties.
Revenue Recognition Notice
If this document is in private pre-General Availability status:
The information contained in this document is for informational sharing purposes only and should be considered in your capacity as a customer advisory board member or pursuant
to your pre-General Availability trial agreement only. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing
decisions. The development, release, and timing of any features or functionality described in this document remains at the sole discretion of Oracle.
This document in any form, software or printed matter, contains proprietary information that is the exclusive property of Oracle. Your access to and use of this confidential material
is subject to the terms and conditions of your Oracle Master Agreement, Oracle License and Services Agreement, Oracle PartnerNetwork Agreement, Oracle distribution agreement,
or other license agreement which has been executed by you and Oracle and with which you agree to comply. This document and information contained herein may not be disclosed,
copied, reproduced, or distributed to anyone outside Oracle without prior written consent of Oracle. This document is not part of your license agreement nor can it be incorporated
into any contractual agreement with Oracle or its subsidiaries or affiliates.
Documentation Accessibility

For information about Oracle's commitment to accessibility, visit the Oracle Accessibility Program website at http://www.oracle.com/pls/topic/lookup?ctx=acc&id=docacc.
Access to Oracle Support

Oracle customers that have purchased support have access to electronic support through My Oracle Support. For information, visit http://www.oracle.com/pls/topic/lookup?
ctx=acc&id=info or visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs if you are hearing impaired.
Référence: E61016
Copyright © 2004, 2020, Oracle et/ou ses affiliés.
Restrictions de licence/Avis d'exclusion de responsabilité en cas de dommage indirect et/ou consécutif
Ce logiciel et la documentation qui l'accompagne sont protégés par les lois sur la propriété intellectuelle. Ils sont concédés sous licence et soumis à des restrictions d'utilisation et
de divulgation. Sauf stipulation expresse de votre contrat de licence ou de la loi, vous ne pouvez pas copier, reproduire, traduire, diffuser, modifier, accorder de licence, transmettre,
distribuer, exposer, exécuter, publier ou afficher le logiciel, même partiellement, sous quelque forme et par quelque procédé que ce soit. Par ailleurs, il est interdit de procéder à toute
ingénierie inverse du logiciel, de le désassembler ou de le décompiler, excepté à des fins d'interopérabilité avec des logiciels tiers ou tel que prescrit par la loi.
Exonération de garantie
Les informations fournies dans ce document sont susceptibles de modification sans préavis. Par ailleurs, Oracle Corporation ne garantit pas qu'elles soient exemptes d'erreurs et vous
invite, le cas échéant, à lui en faire part par écrit.
Avis sur la limitation des droits
Si ce logiciel, ou la documentation qui l'accompagne, est livré sous licence au Gouvernement des Etats-Unis, ou à quiconque qui aurait souscrit la licence de ce logiciel pour le
compte du Gouvernement des Etats-Unis, la notice suivante s'applique :
U.S. GOVERNMENT END USERS: Oracle programs (including any operating system, integrated software, any programs embedded, installed or activated on delivered hardware,
and modifications of such programs) and Oracle computer documentation or other Oracle data delivered to or accessed by U.S. Government end users are "commercial computer
software" or "commercial computer software documentation" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental regulations. As such, the
use, reproduction, duplication, release, display, disclosure, modification, preparation of derivative works, and/or adaptation of i) Oracle programs (including any operating system,
integrated software, any programs embedded, installed or activated on delivered hardware, and modifications of such programs), ii) Oracle computer documentation and/or iii) other
Oracle data, is subject to the rights and limitations specified in the license contained in the applicable contract. The terms governing the U.S. Government's use of Oracle cloud
services are defined by the applicable contract for such services. No other rights are granted to the U.S. Government.
Avis sur les applications dangereuses
Ce logiciel ou matériel a été développé pour un usage général dans le cadre d'applications de gestion des informations. Ce logiciel ou matériel n'est pas conçu ni n'est destiné à
être utilisé dans des applications à risque, notamment dans des applications pouvant causer un risque de dommages corporels. Si vous utilisez ce logiciel ou matériel dans le cadre
d'applications dangereuses, il est de votre responsabilité de prendre toutes les mesures de secours, de sauvegarde, de redondance et autres mesures nécessaires à son utilisation dans
des conditions optimales de sécurité. Oracle Corporation et ses affiliés déclinent toute responsabilité quant aux dommages causés par l'utilisation de ce logiciel ou matériel pour des
applications dangereuses.
Marques
Oracle et Java sont des marques déposées d'Oracle Corporation et/ou de ses affiliés. Tout autre nom mentionné peut correspondre à des marques appartenant à d'autres propriétaires
qu'Oracle.
Intel et Intel Inside sont des marques ou des marques déposées d'Intel Corporation. Toutes les marques SPARC sont utilisées sous licence et sont des marques ou des marques
déposées de SPARC International, Inc. AMD, Epyc, et le logo AMD sont des marques ou des marques déposées d'Advanced Micro Devices. UNIX est une marque déposée de The
Open Group.
Avis d'exclusion de responsabilité concernant les services, produits et contenu tiers
Ce logiciel ou matériel et la documentation qui l'accompagne peuvent fournir des informations ou des liens donnant accès à des contenus, des produits et des services émanant de
tiers. Oracle Corporation et ses affiliés déclinent toute responsabilité ou garantie expresse quant aux contenus, produits ou services émanant de tiers, sauf mention contraire stipulée
dans un contrat entre vous et Oracle. En aucun cas, Oracle Corporation et ses affiliés ne sauraient être tenus pour responsables des pertes subies, des coûts occasionnés ou des
dommages causés par l'accès à des contenus, produits ou services tiers, ou à leur utilisation, sauf mention contraire stipulée dans un contrat entre vous et Oracle.
Date de publication et mention de la version préliminaire de Disponibilité Générale ("Pre-GA")
Version préliminaire de Disponibilité Générale ("Pre-GA") : 15.01.2020
Avis sur la version préliminaire de Disponibilité Générale ("Pre-GA") de la documentation
Si ce document est fourni dans la Version préliminaire de Disponibilité Générale ("Pre-GA") à caractère public ou privé :
Cette documentation est fournie dans la Version préliminaire de Disponibilité Générale ("Pre-GA") et uniquement à des fins de démonstration et d'usage à titre préliminaire de la
version finale. Celle-ci n'est pas toujours spécifique du matériel informatique sur lequel vous utilisez ce logiciel. Oracle Corporation et ses affiliés déclinent expressément toute
responsabilité ou garantie expresse quant au contenu de cette documentation. Oracle Corporation et ses affiliés ne sauraient en aucun cas être tenus pour responsables des pertes
subies, des coûts occasionnés ou des dommages causés par l'utilisation de cette documentation.
Mention sur les informations confidentielles Oracle
INFORMATIONS CONFIDENTIELLES ORACLE. Destinées uniquement à un usage autorisé. Ne pas distribuer à des tiers.
Avis sur la reconnaissance du revenu
Si ce document est fourni dans la Version préliminaire de Disponibilité Générale ("Pre-GA") à caractère privé :
Les informations contenues dans ce document sont fournies à titre informatif uniquement et doivent être prises en compte en votre qualité de membre du customer advisory board ou
conformément à votre contrat d'essai de Version préliminaire de Disponibilité Générale ("Pre-GA") uniquement. Ce document ne constitue en aucun cas un engagement à fournir des
composants, du code ou des fonctionnalités et ne doit pas être retenu comme base d'une quelconque décision d'achat. Le développement, la commercialisation et la mise à disposition
des fonctions ou fonctionnalités décrites restent à la seule discrétion d'Oracle.
Ce document contient des informations qui sont la propriété exclusive d'Oracle, qu'il s'agisse de la version électronique ou imprimée. Votre accès à ce contenu confidentiel et son
utilisation sont soumis aux termes de vos contrats, Contrat-Cadre Oracle (OMA), Contrat de Licence et de Services Oracle (OLSA), Contrat Réseau Partenaires Oracle (OPN),
contrat de distribution Oracle ou de tout autre contrat de licence en vigueur que vous avez signé et que vous vous engagez à respecter. Ce document et son contenu ne peuvent en
aucun cas être communiqués, copiés, reproduits ou distribués à une personne extérieure à Oracle sans le consentement écrit d'Oracle. Ce document ne fait pas partie de votre contrat
de licence. Par ailleurs, il ne peut être intégré à aucun accord contractuel avec Oracle ou ses filiales ou ses affiliés.
Accessibilité de la documentation

Pour plus d'informations sur l'engagement d'Oracle pour l'accessibilité de la documentation, visitez le site Web Oracle Accessibility Program, à l'adresse : http://www.oracle.com/
pls/topic/lookup?ctx=acc&id=docacc.
Accès aux services de support Oracle

Les clients Oracle qui ont souscrit un contrat de support ont accès au support électronique via My Oracle Support. Pour plus d'informations, visitez le site http://www.oracle.com/
pls/topic/lookup?ctx=acc&id=info ou le site http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs si vous êtes malentendant.
Contents

Using This Documentation .................................................................................  9

1 About Managing File Systems ......................................................................  11

About File Systems .......................................................................................  11
Supported File Systems ..................................................................................  12
Disk-Based File Systems ........................................................................  12
Network-Based File Systems ...................................................................  13
Virtual File Systems ............................................................................... 13
Extended File Attributes .................................................................................  15

2 Managing File Systems .................................................................................  17

Creating a File System ...................................................................................  17
Creating a UFS File System ....................................................................  18
Creating a Multi-Terabyte UFS File System ...............................................  19
Customizing UFS File System Parameters .................................................  19
Creating a PCFS File System ..................................................................  22
Creating a File System for CD or DVD .....................................................  23
Determining the Type of a File System .............................................................  24
Mounting a File System .................................................................................  25
Mounting a Disk Based File System .........................................................  25
Mounting a PCFS File System ................................................................. 26
Mounting a Network Based File System ....................................................  27
Mounting a File System Image ................................................................  28
The Virtual File System Table .........................................................................  28
Field Descriptions for the /etc/vfstab File ...............................................  29
Displaying the Mounted File System ........................................................  31
Unmounting a File System ..............................................................................  33
▼ How to Stop All Processes That Are Accessing a File System ...................  34

7
Contents

Checking UFS File System Consistencies ..........................................................  35

▼ How to Check a UFS File System .......................................................  35
Preening UFS File Systems .....................................................................  37
Resolving UFS File System Inconsistencies ...............................................  38
Fixing a UFS File System That the fsck Command Cannot Repair .................  59
▼ How to Restore a Bad Superblock .......................................................  60

3 Managing Data and Free Space ....................................................................  63

File System Space Usage ................................................................................  63
Files Using the Most Space in a File System ......................................................  65
Quotas in File System ....................................................................................  66

4 File System Backup and Restore ..................................................................  69

Backing Up and Restoring UFS File Systems .....................................................  69
How the ufsdump Command Works ..........................................................  69
Copying Data Using the ufsdump Command ..............................................  70
Specifying Files to Back Up .................................................................... 72
Specifying Tape Characteristics ................................................................ 72
Limitations of the ufsdump Command .......................................................  73
Specifying the ufsdump Command Options and Arguments ...................................  73
Default ufsdump Options ........................................................................  73
The ufsdump Command and Security Issues ...............................................  74
Specifying ufsrestore Options and Arguments ..................................................  74
Copying Files and File System ........................................................................  75
Commands for Copying File Systems .......................................................  75
Copying Directories Between File Systems Using cpio Command .................. 76
Copying Files and File Systems to a Tape Device ........................................ 77
Copying Files and File Systems to a Remote Tape Device ............................. 86

5 Managing File System Performance .............................................................  89

Monitoring File System Performance Using fsstat ............................................. 89
Tuning UFS File System Performance Using tunefs ...........................................  90
UFS Direct I/O Mode ....................................................................................  90

Index .................................................................................................................. 91

8 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Using This Documentation

■ Overview – Provides an overview of Oracle Solaris file systems, including information

about how to manage them and how to perform file system administration tasks.
■ Audience – System administrators.
■ Required knowledge – Basic Oracle Solaris or UNIX system administration experience
and general file system administration experience.

Product Documentation Library

Documentation and resources for this product and related products are available at http://www.
oracle.com/pls/topic/lookup?ctx=E37838-01.

Feedback
Provide feedback about this documentation at http://www.oracle.com/goto/docfeedback.

Using This Documentation 9

10 Managing File Systems in Oracle Solaris 11.4 • November 2020
 ♦ ♦ ♦
1
C H A P T E R 1

About Managing File Systems

Managing file systems, data, free space, file system backup, and restore are some of the most
important system administration tasks. This chapter provides an overview of how files and
directories are organized to form a file system, and describes different types of file system
supported by Oracle Solaris.
This chapter contains the following topics:
■ “About File Systems” on page 11
■ “Supported File Systems” on page 12
■ “Extended File Attributes” on page 15

About File Systems

It is common to have all the files organized in a single structure of directories, called as a file
system tree in a UNIX environment. The directory structure starts with a root directory (/).

Files can be stored on different types of media or networks. The UNIX system enables you to
attach and detach files stored on a given media. The process of attaching and detaching files
from the directory is called mounting and unmounting of files. The directory, where a subtree or
a file from one media is attached to the complete directory structure, is called a mount point.

A file system is a structure of directories that is used to organize and store files. Any of the
items from the following list can be termed as a file system:
■ Particular type of file system: disk-based, network-based, or virtual
■ Entire file system tree, beginning with the root (/) directory
■ Data structure of a disk slice or other media storage device
■ Portion of a file tree structure that is attached to a mount point on the main file tree so that
the files are accessible

The Oracle Solaris operating system (OS) uses the virtual file system (VFS) architecture, which
provides a common interface for various file system types. The VFS architecture enables the

Chapter 1 • About Managing File Systems 11

Supported File Systems

kernel to provide a common interface for usual operations such as reading, writing, and listing
files. The execution of these operations are managed by kernel modules for a specific file
system. The VFS architecture also makes it easier to add new file systems.

Supported File Systems

The Oracle Solaris OS supports file systems stored on a locally attached storage media such as,
a hard disk and file systems stored on a network. Storing file systems locally or on a network
enables the VFS architecture to provide information from kernel.

Disk-Based File Systems

Disk-based file systems enable you to access files stored on a locally attached media such as
hard disks, CD-ROMs, or USB devices. Oracle Solaris supports different types of disk-based
file systems.

UFS Legacy UNIX file system (based on the BSD Fat Fast File system).

ZFS Zettabyte file system (ZFS) is the default disk-based and root file system.
The ZFS file system has some features which are not found in any other
file system, such as, creating storage pools, snapshots, and using copy-on
write semantics. These features are not available in any other file system.
For more information about ZFS file system, see Managing ZFS File
Systems in Oracle Solaris 11.4.

HSFS High Sierra File System (HSFS) is used on CD-ROMs and DVDs. Oracle
Solaris supports file systems created according to ECMA-119, ISO 9660:
1998 or ISO 9660:1999 with Rock Ridge and Joliet extensions. The
extensions provide additional features and attributes to files such as long
file names (up to 255 characters), and UNIX permissions and attributes.
Due to the nature of the media, a HSFS file system is read only.

PCFS PC file system (PCFS), which enables read and write access to data
and programs on DOS-formatted disks that are written for DOS-based
personal computers.

UDFS The Universal Disk Format (UDFS) file system which is the industry-
standard format for storing information on the optical media technology
called DVD (Digital Versatile Disc or Digital Video Disc).

12 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Supported File Systems

SAM-QFS SAM-QFS is an integrated hierarchical storage manager (HSM) and

storage area network (SAN) file system. SAM is a component of
HSM storage and archive management. QFS is the SAN scalable high
performance file system component. SAM-QFS also has an integrated
disk volume management and tape volume management. QFS also has
a write once, read many times (WORM) file system capability. QFS can
be used independently of SAM when just a file system is needed. SAM
requires QFS and cannot be used independently of QFS.

Network-Based File Systems

Network-based file systems can be accessed from a network. Typically, network-based file
systems reside on a server, and are accessed by other systems across the network. The different
types of network-based file systems are as follows:

NFS Network File System (NFS) is a common resource sharing service in an

UNIX environment. With the NFS service, you can provide distributed
resources (files or directories) by sharing them from a server and
mounting them on multiple clients. Oracle Solaris support versions 2, 3
and 4 of the NFS protocol. For more information, see Managing Network
File Systems in Oracle Solaris 11.4.

SMB or CIFS Server Message Block (SMB) protocol is a resource sharing service
used primarily in workgroups or domains of Microsoft Windows
systems. With the Oracle SMB service, you can provide distributed
resources (files or directories) to Windows and Mac OS systems by
sharing them from a server and mounting them on multiple clients.
For more information, see Managing SMB File Sharing and Windows
Interoperability in Oracle Solaris 11.4.

Virtual File Systems

Virtual file systems are memory-based file systems that provide access to special kernel
information and facilities. Most virtual file systems do not use file system disk space. Also,
some virtual file systems, such as the temporary file system (TMPFS) use the swap space on a
disk. The different types of virtual file system are as follows:

CTFS The contract file system (CTFS) is an interface for creating, controlling,
and observing contracts. A contract enhances the relationship between

Chapter 1 • About Managing File Systems 13

Supported File Systems

a process and a system resource. It provides richer error reporting and

means of delaying the removal of a resource.
The service management facility (SMF) uses process contracts (a type of
contract) to track the processes, which compose a service. Using process
contracts helps SMF to identify service failure even if a part of a multi-
process service fails.

MNTFS The mnttab file system (MNTFS) provides read-only access to the table
of mounted file systems for the local system.

OBJFS The object file system (OBJFS) describes the state of all modules
currently loaded by the kernel. This file system is used by debuggers to
access information about kernel symbols without accessing the kernel
directly.

SHAREFS The sharetab file system (SHAREFS) provides read-only access to the
table of shared file systems for the local system.

PROCFS The process file system (PROCFS) resides in memory and contains
a list of active processes, by process number, in the /proc directory.
Information in the /proc directory is used by commands such as ps.
Debuggers and other development tools can also access the address space
of the processes by using the PROCFS calls.

TMPFS The temporary file system (TMPFS) uses local memory for a file system
to read and write. Using TMPFS improves system performance by saving
the cost of reading and writing temporary files to a local disk or across
the network. For example, temporary files are created when you compile
a program. The OS generates disk activity or network activity logs
while manipulating these files. Using TMPFS to hold these temporary
files significantly increases the speed of create, delete, and data access
operations.
Files in TMPFS file systems are not permanent. These files are deleted
when the file system is unmounted or when the system is shut down or
rebooted.
TMPFS is the default file system type for the /tmp directory in the Oracle
Solaris OS. You can copy or move files into or out of the /tmp directory,
as you would in a ZFS or UFS file system. The TMPFS file system uses
swap space as a temporary backing store.

LOFS The loopback file system (LOFS) enables you create a new virtual file
system so that you can access files by using an alternative path name. For

14 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Extended File Attributes

example, you can create a loopback mount of the root (/) directory on
/tmp/newroot. This loopback mount and the entire file system hierarchy
appears as if it is duplicated under /tmp/newroot, including any file
systems mounted from NFS servers. All files are accessible either with
a path name starting from root (/), or with a path name that starts from
/tmp/newroot.

Extended File Attributes

The ZFS, UFS, NFS, and TMPFS file systems are enhanced to include extended file attributes.
Extended file attributes enable application developers to associate specific attributes to a file.
For example, a developer of an application used to manage a windowing system might choose
to associate a display icon with a file. Extended file attributes are logically represented as files
within a hidden directory that are associated with the target file.

You can use the runat command to add attributes and execute shell commands in the extended
attribute namespace. This namespace is a hidden attribute directory that is associated with a
specified file.

To add attributes to a file, you must create the attributes file using the runat command.

$ runat filea cp /tmp/attrdata attr.1

Then, use the runat command to list the attributes of the file.

$ runat filea ls -l

For more information, see the runat(1) man page.

Many Oracle Solaris file system commands are modified to support extended file attributes,
which enables you to query, copy, or find file attributes. For more information, see the specific
man page for each file system command.

Chapter 1 • About Managing File Systems 15

16 Managing File Systems in Oracle Solaris 11.4 • November 2020
 ♦ ♦ ♦
2
C H A P T E R 2

Managing File Systems

This chapter describes how to create, mount, and unmount a file system. This chapter also
describes how to resolve UFS file system inconsistencies and errors.
This chapter contains the following topics:

■ “Creating a File System” on page 17

■ “Determining the Type of a File System” on page 24
■ “Mounting a File System” on page 25
■ “The Virtual File System Table” on page 28
■ “Unmounting a File System” on page 33
■ “Checking UFS File System Consistencies” on page 35
■ “Resolving UFS File System Inconsistencies” on page 38

Creating a File System

To create a new file system, you must format and label a disk or a storage device such as USB,
flash disk, or memory card. You can also create a file system spanning a whole device. Usually
volumes are provided by a volume manager such as Oracle Solaris volume manager (SVM).
There are no procedures to create network-based and virtual file systems.

For information about how to create a ZFS file system, see Managing ZFS File Systems in
Oracle Solaris 11.4.

Caution - While creating a new file system, ensure that you specify the correct device name. If
you specify an incorrect device name, the content in that device will be erased. This error might
cause the system to panic.

You can use the mkfs command to create a file system.

# mkfs -F type [-o parameters] raw_device [parameters]

Chapter 2 • Managing File Systems 17

Creating a File System

Parameters differ based on the type of file system you want to create. The syntax can also differ
from one file system to the other, though there are some common parameters such as block-
size. For more information, see the mkfs(8) man page.

Creating a UFS File System

The mkfs tool requires a number of parameters related to device geometry for UFS to be
entered. The newfs tool obtains these parameters automatically.

# newfs [-N] [-b size] [-i bytes] /dev/rdsk/device-name

The -b option specifies the size of a block (a unit in which the data is transferred to and from the
device). The default size of a block is 8192 bytes.

The -i option specifies the number of bytes per inode, which specifies the density of inodes in
the file system. The value should reflect the expected average size of files in the file system. If
you require fewer inodes, larger number must be used. To create more inodes, you must provide
a smaller number.

When you issue the newfs command, it prompts for confirmation if the file system must
be created on a given device. The newfs tool then invokes the mkfs tool with appropriate
arguments for the device.

EXAMPLE 1 Creating a UFS File System

The following example shows how to create a UFS file system on /dev/rdsk/c0t1d0s0.

# newfs /dev/rdsk/c0t1d0s0
newfs: construct a new file system /dev/rdsk/c0t1d0s0: (y/n)? y
/dev/rdsk/c0t1d0s0: 286722656 sectors in 46668 cylinders of 48 tracks, 128 sectors
140001.3MB in 2917 cyl groups (16 c/g, 48.00MB/g, 5824 i/g)
super-block backups (for fsck -F ufs -o b=#) at:
32, 98464, 196896, 295328, 393760, 492192, 590624, 689056, 787488, 885920,
Initializing cylinder groups:
..........................................................
super-block backups for last 10 cylinder groups at:
285773216, 285871648, 285970080, 286068512, 286166944, 286265376, 286363808,
286462240, 286560672, 286659104

For more information, see the newfs(8)man page.

18 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Creating a File System

Creating a Multi-Terabyte UFS File System

The UFS file system size is limited to 2 TB. In some environments such as a database storage,
you need to store more data in very few files. UFS provides a configuration called Multi-
TeraByte UFS (or MTBUFS), which can store up to 16 TB data.

To create a multi-terabyte UFS, run the newfs command with the -T option.

# newfs -T raw_device

Assuming that the file system might contain only a small number of files, the space needed for
metadata (information about the files) is reduced and reused for additional data capacity.

EXAMPLE 2 Creating a Multi-Terabyte UFS File System

The following example shows how to create a MTBUFS file system on /dev/rdsk/c2t2d0s0.

# newfs -T /dev/rdsk/c2t2d0s0
newfs: construct a new file system /dev/rdsk/c2t2d0s0: (y/n)? y
Warning: 2864 sector(s) in last cylinder unallocated
/dev/rdsk/c2t2d0s0: 143299792 sectors in 23324 cylinders of 48 tracks, 128 sectors
69970.6MB in 1458 cyl groups (16 c/g, 48.00MB/g, 64 i/g)
super-block backups (for fsck -F ufs -o b=#) at:
32, 98464, 196896, 295328, 393760, 492192, 590624, 689056, 787488, 885920,
Initializing cylinder groups:
............................
super-block backups for last 10 cylinder groups at:
142349344, 142447776, 142546208, 142644640, 142743072, 142841504, 142939936,
143038368, 143136800, 143235232

Customizing UFS File System Parameters

Ensure that you understand the available parameters for the newfs command before you modify
the default file system parameters. For more information about customizing these parameters,
see the newfs(8) and mkfs_ufs(8) man pages.

Logical Block Size

The logical block size parameter can be set by using the -b bsize parameter with the newfs
command.

Chapter 2 • Managing File Systems 19

Creating a File System

The logical block size is the size of the block that the UNIX kernel uses to read or write a file.
The logical block size is usually different from the physical block size. The physical block size
is usually 512 bytes, which is the size of the smallest block that a disk controller can read or
write. Logical block size is set to the page size of the system by default. The default logical
block size is 8192 bytes (8 KB) for UFS file systems. The UFS file system supports block sizes
of 4096 or 8192 bytes (4 or 8 KB). It is a best practice to set the logical block size to 8 KB.

To choose the best logical block size suitable for your system, you must consider both
performance and available space. For most UFS systems, an 8 KB file system provides the best
performance, offering a good balance between disk performance, and the use of space in the
primary memory and on the disk.

To increase efficiency, you can use a larger logical block size for file systems when most of the
files are large. Use a smaller logical block size for file systems when most of the files are small.
You can use the -c option with the newfs command on a file system to display a complete report
on the distribution of files by block size. However, the page size that is set when the file system
is created is probably the best size in most cases.

Fragment Size

The fragment size parameter can be set using the -f fragsize parameter with the newfs
command.

As files are created or expanded, they are allocated with disk space with either full logical
blocks or portions of logical blocks called fragments. When disk space is needed for a file,
full blocks are allocated first, and then one or more fragments of a block are allocated for the
remainder. For small files, allocation begins with fragments.

The ability to allocate fragments of blocks to files saves space by reducing fragmentation of
disk space that result from unused holes in blocks.

You define the fragment size when you create a UFS file system. The default fragment size is 1
KB. Each block can be divided into 1, 2, 4, or 8 fragments, which results in fragment sizes from
8192 bytes to 512 bytes (for 4 KB file systems only). The lower bound is actually tied to the
disk sector size, typically 512 bytes.

For multi-terabyte file systems, the fragment size must be equal to the file system block size.

When choosing a fragment size, consider the trade-off between space and time. A small
fragment size saves space, but requires more time to allocate. To increase storage efficiency, use
a larger fragment size for file systems when most of the files are large. Use a smaller fragment
size for file systems when most of the files are small.

20 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Creating a File System

Minimum Free Space

The minimum free space is the percentage of the total disk space that is held in reserve when
you create a file system. The default reserve is ((64 MB/partition size) * 100), rounded down to
the nearest integer and limited between 1 percent and 10 percent, inclusively.

Free space is important because file access becomes less efficient when a file system becomes
full. As long as an adequate amount of free space exists, UFS file systems operate efficiently.
When a file system becomes full, using the available user space, only root can access the
reserved free space.

Commands such as df reports the percentage of space that is available to users, excluding the
percentage allocated as the minimum free space. When the command reports that more than 100
percent of the disk space in the file system is in use, some of the reserve space has been used by
root.

If you impose quotas on users, the amount of space available to them does not include the
reserved free space. You can change the value of the minimum free space for an existing file
system using the tunefs command.

Optimization Type
The optimization type (-o option) is set to either space or time.

space When you select space optimization, disk blocks are allocated to
minimize fragmentation and disk use is optimized.

time When you select time optimization, disk blocks are allocated as quickly
as possible, with less emphasis on their placement. When sufficient free
space exists, allocating disk blocks is relatively easy, without resulting in
too much fragmentation. The default is time.

You can change the value of the optimization type parameter for an existing file system by
using the tunefs command. For more information, see the tunefs(8) man page.

Number of Inodes
The number of bytes per inode specifies the density of inodes in the file system. The number
of bytes is divided into the total size of the file system to determine the number of inodes to
create. Once the inodes are allocated, you cannot change the number without re-creating the file
system.

Chapter 2 • Managing File Systems 21

Creating a File System

If the file system is less than 1 GB, the default number of bytes per inode is 2048 bytes (2 KB).
The following table shows the number of bytes per inode for particular file system size:

File System Size Number of Bytes Per Inode

Less than or equal to 1 GB 2048
Less than 2 GB 4096
Less than 3 GB 6144
3 GB up to 1 TB 8192
Greater than 1 TB or created with the -T option 1048576

If you have a file system with many symbolic links, they can lower the average file size. If the
file system has many small files, provide a lower value for the inode parameter.

Caution - If you have less number of inodes, you reach the maximum number of files on
a disk slice that is practically empty. Hence, ensure that the number of inodes match your
requirements.

Maximum UFS File and File System Size

The maximum size of a UFS file system is about 16 TB of usable space, minus about one
percent overhead. A sparse file can have a logical size of one terabyte. However, the actual
amount of data that can be stored in a file is approximately one percent less than 1 TB because
of the file system overhead.

Maximum Number of UFS Subdirectories

The maximum number of subdirectories per directory in a UFS file system is 32,767. This limit
is predefined and cannot be changed.

Creating a PCFS File System

To create a PCFS (or FAT) file system, use the mkfs command. However, PCFS requires special
care for specifying the device and might require some additional parameters. The syntax for the
mkfs command is as follows:

# mkfs -F pcfs [-o parameters] device-name[:logical-name] [size]

22 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Creating a File System

As PCFS is usually placed on a DOS partition, Oracle Solaris uses a special naming scheme to
access these partitions. The device specification consists of two parts:

/dev/rdsk/ Specifies the device name of the whole disk. For example, /dev/rdsk/
device-name c0t0d0p0 on an x86 system or /dev/rdsk/c0t0d0s2 on a SPARC
system.

logical-drive Specifies either the DOS logical drive letter (C through Z) or a drive
number (1 through 24). Drive C is equivalent to drive 1 and represents
the primary DOS partition on the drive. All other letters or numbers
represent DOS logical drives within the extended DOS partition.

Alternatively, the partition might be specified as /dev/rdsk/c0t0d0pX on an x86 system, where

X is 1 to 4 for primary partitions and 5 to 32 for logical partitions. However, mkfs requires
additional argument nofdisk. This parameter has to also be used when the PCFS is created on a
slice or on a whole device.

Caution - Ensure that the partition does not cover the sectors containing the label information.
PCFS does not protect the first sectors of a partition or slice (unlike UFS and ZFS), and
overwrites them with its own data. Creating PCFS file system on such a partition or slice
destroys the label and all data on the device is lost.

Creating a File System for CD or DVD

There are several file systems that can be used on CD or DVD media. Oracle Solaris supports
HSFS (High Sierra File System also known as ISO-9660) and UDF (Universal Disk Format)
file system. mkisofs is a pre-mastering tool used to create pure or hybrid image of HSFS/
UDF/HFS file systems with an optional Rock-Ridge extension. A hybrid image is an effective
combination of selected file systems containing information for all of them. Such an image can
be accessed as any of the reselected ones. The syntax for the mkisofs command is as follows:

# mkisofs [options] [-o image-name] path [path ...]

mkisofs creates an image containing all files specified by paths, and stores it to a file specified
by image-name (or the standard output if none is given). The image can then be burned to a CD
or DVD media using cdrecord or a similar tool.

As mkisofs is a third party tool, certain combinations of parameters might result in an

unreadable image by Oracle Solaris.

Chapter 2 • Managing File Systems 23

Determining the Type of a File System

EXAMPLE 3 Creating an HSFS Image Using the mkisofs Command

The following example shows how to create an HSFS compliant image with Rock-Ridge
extension using the mkisofs command. The content is specified by a path to a directory.

# mkisofs -R -J -o /data/hsfsimage.iso /data/content

0.12% done, estimate finish Mon Sep 12 15:01:35 2016
0.24% done, estimate finish Mon Sep 12 14:47:33 2016
...
49.97% done, estimate finish Mon Sep 12 14:43:05 2016
50.09% done, estimate finish Mon Sep 12 14:43:04 2016
50.21% done, estimate finish Mon Sep 12 14:43:04 2016
...
99.82% done, estimate finish Mon Sep 12 14:43:19 2016
99.94% done, estimate finish Mon Sep 12 14:43:19 2016
Total translation table size: 0
Total rockridge attributes bytes: 2175
Total directory bytes: 0
Path table size(bytes): 10
Max brk space used 14000
4222681 extents written (8247 MB)

Determining the Type of a File System

If you have a raw device name of a disk slice, you can use the fstyp or the df command to
determine the type of a file system. The fstyp command might not be able to identify third
party file systems, which are not supported by Oracle Solaris.

For more information, see the fstyp(8) and df(8) man pages.

EXAMPLE 4 How to Determine the Type of a File System

The following example uses the df -n command to display the mounted file systems along with
its type. The df command uses information about mounted file systems provided by the kernel.

# df -n
/ : zfs
/devices : devfs
/dev : dev
/system/contract : ctfs
/proc : proc
/etc/mnttab : mntfs
/system/volatile : tmpfs

24 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Mounting a File System

/system/object : objfs
/etc/dfs/sharetab : sharefs
/dev/fd : fd
/var : zfs
/tmp : tmpfs
/var/share : zfs
/export : zfs
/export/home : zfs
/rpool : zfs
/media/cdrom : ufs
/media/cdrom-1 : ufs
/media/cdrom-2 : ufs
/media/cdrom-3 : ufs
/media/sol_10_811_sparc : hsfs
/media/cdrom-4 : ufs
/pond : zfs
/pond/amy : zfs
/pond/dr : zfs
/pond/rory : zfs

The following example uses the fstyp command to determine the file system type. The
specified device might or might not be mounted.
# fstyp /dev/rdsk/c0t0d0s0
zfs

Mounting a File System

Mounting operation connects the file system subtree that is on a specific device to the file
system tree. The transition between different file systems is transparent and is equal to
navigating to another directory.

To mount a file system, use the mount command. For more information, see the mount(8) man
page.

For information about mounting ZFS file systems, see “Mounting ZFS File Systems” in
Managing ZFS File Systems in Oracle Solaris 11.4.

Mounting a Disk Based File System

Apart from the device specification, all the command arguments are similar to other file
systems. For disk based file system, a path to block device specifies the device. The syntax for
the mount command is as follows:

Chapter 2 • Managing File Systems 25

Mounting a File System

# mount [-F fs-type] [-o mount-options] /dev/dsk/device-name /directory-name

-F fs-type The type of file system to be mounted. If the option is not set, mount
assumes that it is a UFS file system.

-o mount-options Some mount options are common to all types of file system. However,
each file system provides a different list of mount options depending on
its features. Some of the common options are rw and ro, which set the
access to the file system to read and write, or read-only.

/dev/dsk/device- The path to the block device of a slice or partition containing the file
name system.

/directory-name The directory where you want to mount the file system.

When a file system is specified in the virtual file system table /etc/vfstab, you can omit the
device name or the mount point. The mount command refers to the /etc/vfstab file for missing
information and mount options, and merges them with the arguments provided by the user.

EXAMPLE 5 Mounting a UFS File System

The following example shows how to mount a UFS in slice 3 to the /export/park directory.

# mount -F ufs /dev/dsk/c0t0d0s3 /export/park

Mounting a PCFS File System

Oracle Solaris uses a special naming scheme to access DOS partitions where the PCFS file
system is usually stored. The resource consists of a path to a block device of the whole disk and
a logical drive specifier.

# mount -F pcfs /dev/dsk/device-name:logical-drive /mount-point

/dev/dsk/device- Specifies the device name of the whole disk. For example, /dev/dsk/
name c0t0d0p0 on an x86 system or /dev/dsk/c0t0d0s2 on a SPARC system.

logical-drive Specifies either the DOS logical drive letter (C through Z) or a drive
number (1 through 24). Drive C is equivalent to drive 1, and represents
the primary DOS partition on the drive. All other letters or numbers
represent DOS logical drives within the extended DOS partition.

Note - You must use a colon to separate the device-name and logical-drive.

26 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Mounting a File System

EXAMPLE 6 Mounting a PCFS File System From a Primary Partition on x86 and SPARC Systems

The following example shows how to mount a logical drive in the primary DOS partition on the
/pcfs/c directory on an x86 system.

# mount -F pcfs /dev/dsk/c0t0d0p0:c /pcfs/c

The following example shows how to mount a logical drive in the primary DOS partition on the
/pcfs/c directory on a SPARC system.

# mount -F pcfs /dev/dsk/c0t0d0s2:c /pcfs/c

EXAMPLE 7 Mounting a PCFS (DOS) File System From a Logical Partition on x86 and SPARC
Systems

The following example shows how to read only mount the first logical drive in the extended
DOS partition located on the /mnt directory on an x86 system.

# mount -F pcfs -o ro /dev/dsk/c0t0d0p0:2 /mnt

The following example shows how to read only mount the first logical drive in the extended
DOS partition located on the /mnt directory on a SPARC system.

# mount -F pcfs -o ro /dev/dsk/c0t0d0s2:2 /mnt

Caution - On an x86 system, the DOS partitions are available as devices named c0t0d0p1 to
c0t0d0p4 for primary partitions and c0t0d0p5 to c0t0d0p32 for logical partitions. Those can
also be used to mount and access PCFS file system. Ensure to use notation either consistently or
exclusively. Using the same file system twice can corrupt the file system.

Mounting a Network Based File System

For a network based file system, the resource specification consists of a server name and a path
to a directory. This resource specification is also called a share. For more information about
how to create a share over a specific network protocol see, Managing Network File Systems in
Oracle Solaris 11.4 and Managing SMB File Sharing and Windows Interoperability in Oracle
Solaris 11.4. You can also refer to your operating system documentation.

# mount -F (nfs|smbfs) [-o mount-options] server:/directory /mount-point

You can automate the network based file systems by using the automount command. For more
information about the automounter, see Chapter 4, “Administering Autofs” in Managing
Network File Systems in Oracle Solaris 11.4 and the automount(8) man page.

Chapter 2 • Managing File Systems 27

The Virtual File System Table

EXAMPLE 8 Mounting an NFS File System

The following example shows how to mount the /export/packages directory on /mnt from the
server, pluto.

# mount -F nfs pluto:/export/packages /mnt

EXAMPLE 9 Mounting a CIFS or an SMB File System

The following example shows how to mount the /data/workspace directory on /mnt from the
server, neptune.

# mount -F smbfs neptune:/data/workspace /mnt

Mounting a File System Image

You can specify a regular file containing a file system image instead of specifying a device.

# mount -F fs-type [-o mount-options] path-to-image-file /mount-point

Depending on the capability of the file system driver, an image is mounted as read-only, or
read-write. For example, HSFS image can be mounted only for reading while PCFS image can
be mounted for reading or writing.

EXAMPLE 10 Mounting an Image of HSFS File System

The following example shows how to mount the HSFS image stored in /tank/build/
image.iso on the /mnt file system.

# mount -F hsfs /tank/build/image.iso /mnt

The Virtual File System Table

Most file systems are mounted automatically by an SMF service at the system boot time. All
file systems except for ZFS are mounted according to the /etc/vfstab file. The list of ZFS file
systems to mount is taken from ZFS cache.

To add an entry for mounting a legacy or remote file system, the information you need to
specify are as follows:

28 Managing File Systems in Oracle Solaris 11.4 • November 2020

 The Virtual File System Table

■ The device or the server where the file system resides

■ The file system mount point
■ File system type
■ Whether you want the file system to mount automatically when the system boots
■ Any mount options

The following vfstab example is from a system that has a ZFS root file system. In addition,
this system is mounting a remote file system, /users/data, from the NFS server, neo.

# cat /etc/vfstab
#device device mount FS fsck mount mount
#to mount to fsck point type pass at boot options
#
fd - /dev/fd fd - no -
/proc - /proc proc - no -
/dev/zvol/dsk/rpool/swap - - swap - no -
/devices - /devices devfs - no -
sharefs - /etc/dfs/sharetabsharefs - no -
ctfs - /system/contract ctfs - no -
objfs - /system/object objfs - no -
swap - /tmp tmpfs - yes -
neo:/users/data - /data nfs - yes -

You can mount ZFS file systems from the vfstab file by using the legacy mount feature.

Field Descriptions for the /etc/vfstab File

An entry in the /etc/vfstab file has seven fields:

device to mount Identifies one of the following resources:

■ Block device name for a local legacy UFS file system. For example,
/dev/dsk/c8t1d0s7.
■ Resource name for a remote file system. For example, myserver:/
export/home.
After you add an entry for a remote system resource, ensure that the
following service is enabled.

# svcs -a | grep nfs/client

disabled May_14 svc:/network/nfs/client:default
# svcadm enable svc:/network/nfs/client:default

Chapter 2 • Managing File Systems 29

The Virtual File System Table

■ Swap volume. For example, /dev/zvol/dsk/rpool/swap.

■ Directory for a virtual file system.

device to fsck Determines the raw interface that is used by the fsck command. Use a
dash (-) when there is no applicable device, such as for a read-only file
system or a remote file system. The raw device name that corresponds to
the legacy UFS file system identified by the device to mount field.

mount point Identifies where to mount the legacy or remote file system.

FS type Identifies the type of file system.

fsck pass The pass number indicates whether to check a legacy UFS file system.
When the field contains a dash (-), the file system is not checked. In
Oracle Solaris 11.4, the fsck pass value in the /etc/vfstab file is
ignored during the boot process.
When this field is zero, legacy UFS file systems are not checked. When
the field value is greater than zero, the UFS file system is checked.
All legacy UFS file systems with a value of 1 in this field are checked
one at a time in the order they appear in the vfstab file. When the
fsck command is run on multiple UFS file systems that have fsck
pass values greater than 1, using the preen option, the fsck command
automatically checks the file systems on different disks in parallel to
maximize efficiency. Otherwise, the value of the pass number does not
have any effect.

mount at boot Specifies whether the file system must be automatically mounted by
mountall command when the system is booted. The option can be yes or
no. This field has nothing to do with autofs. This field should always be
set to no for virtual file systems such as /proc and /dev/fd.

mount options A list of comma separated options (with no spaces) that are used for
mounting the file system. Use a dash (-) to indicate no options. For more
information, see the vfstab(5) man page.

Note - You must have an entry in each field in the /etc/vfstab file. If no value is supplied, the
system boot might not be successful. Ensure that you add a dash (-) because white space cannot
be used as a field value.

For more information about NFS, see Managing Network File Systems in Oracle Solaris 11.4.

30 Managing File Systems in Oracle Solaris 11.4 • November 2020

 The Virtual File System Table

Displaying the Mounted File System

You can display the list of mounted file systems. There are several tools to display various
information about all the mounted file systems. For example, the df command displays the
capacity and free space on each file system.

For more information about the mounted file system, you can either check the output of the
mount command or check the content of the /etc/mnttab file.

Using the Mounted File System Table

When you mount or unmount a file system, the /etc/mnttab file is updated with the list
of currently mounted file systems. This file cannot be edited by any user including the
administrator. The following example shows the /etc/mnttab file.

$ more /etc/mnttab
rpool/ROOT/zfsBE / zfs dev=3390002 0
/devices /devices devfs dev=8580000 1337114941
/dev /dev dev dev=85c0000 1337114941
ctfs /system/contract ctfs dev=8680001 1337114941
proc /proc proc dev=8600000 1337114941
mnttab /etc/mnttab mntfs dev=86c0001 1337114941
swap /system/volatile tmpfs xattr,dev=8700001 1337114941
objfs /system/object objfs dev=8740001 1337114941
sharefs /etc/dfs/sharetab sharefs dev=8780001 1337114941
/usr/lib/libc/libc_hwcap2.so.1 /lib/libc.so.1 lofs dev=3390002 13371149
fd /dev/fd fd rw,dev=8880001 1337114969
rpool/ROOT/zfsBE/var /var zfs
rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime,dev=3390003 1337114969
swap /tmp tmpfs xattr,dev=8700002 1337114969
rpool/VARSHARE /var/share zfs
rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime,dev=3390004 133711496

Obtaining the List of Mounted File Systems Using the mount

Command
In addition to mounting file systems, the mount command displays the list of all the mounted
file systems when resource and mount point are not specified. This list can be altered by using
the -p option to obtain a format more suitable for processing by other programs or scripts.

The following example shows the list of mounted file systems using the mount command
without any arguments.

Chapter 2 • Managing File Systems 31

The Virtual File System Table

# mount
/ on rpool/ROOT/zfsBE read/write/setuid/devices/rstchown/dev=45d0002 on Thu Jan 1
01:00:00 1970
/devices on /devices read/write/setuid/devices/rstchown/dev=8b00000 on Mon Jun 20
14:46:48 2016
/dev on /dev read/write/setuid/devices/rstchown/dev=8b40000 on Mon Jun 20 14:46:48 2016
/system/contract on ctfs read/write/setuid/devices/rstchown/dev=8c40001 on Mon Jun 20
14:46:48 2016
/proc on proc read/write/setuid/devices/rstchown/dev=8b80000 on Mon Jun 20 14:46:48 2016
/etc/mnttab on mnttab read/write/setuid/devices/rstchown/dev=8c80001 on Mon Jun 20
14:46:48 2016
/system/volatile on swap read/write/setuid/devices/rstchown/xattr/dev=8cc0001 on Mon Jun
20 14:46:48 2016
/system/object on objfs read/write/setuid/devices/rstchown/dev=8d00001 on Mon Jun 20
14:46:48 2016
/etc/dfs/sharetab on sharefs read/write/setuid/devices/rstchown/dev=8d40001 on Mon Jun
20 14:46:48 2016
/lib/libc.so.1 on /usr/lib/libc/libc_hwcap2.so.1 read/write/setuid/devices/rstchown/
dev=45d0002 on Mon Jun 20 14:47:03 2016
/dev/fd on fd read/write/setuid/devices/rstchown/dev=8e00001 on Mon Jun 20 14:47:04 2016
/var on rpool/ROOT/s114_14/var read/write/setuid/devices/rstchown/nonbmand/exec/xattr/
atime/dev=45d0004 on Mon Jun 20 14:47:36 2016
/tmp on swap read/write/setuid/devices/rstchown/xattr/dev=8cc0002 on Mon Jun 20 14:47:36
2016
/var/share on rpool/VARSHARE read/write/nosetuid/devices/rstchown/nonbmand/noexec/
noxattr/atime/dev=45d0005 on Mon Jun 20 14:47:37 2016
/export on rpool/export read/write/setuid/devices/rstchown/nonbmand/exec/xattr/atime/
dev=45d0006 on Mon Jun 20 14:47:53 2016
/export/home on rpool/export/home read/write/setuid/devices/rstchown/nonbmand/exec/
xattr/atime/dev=45d0007 on Mon Jun 20 14:47:54 2016
/export/home/jack on rpool/export/home/jack read/write/setuid/devices/rstchown/nonbmand/
exec/xattr/atime/dev=45d0008 on Mon Jun 20 14:47:56 2016
/rpool on rpool read/write/setuid/devices/rstchown/nonbmand/exec/xattr/atime/dev=45d0009
on Mon Jun 20 14:47:56 2016
/mnt on /dev/dsk/c2t1d0s0 read/write/setuid/devices/rstchown/intr/largefiles/logging/
xattr/onerror=panic/dev=3400080 on Tue Jun 21 13:39:25 2016

The following example shows the output suitable for automated parsing. The format is the same
as the /etc/vfstab file.

# mount -p
rpool/ROOT/zfsBE - / zfs - no
/devices - /devices devfs - no
/dev - /dev dev - no
ctfs - /system/contract ctfs - no
proc - /proc proc - no

32 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Unmounting a File System

mnttab - /etc/mnttab mntfs - no

swap - /system/volatile tmpfs - no xattr
objfs - /system/object objfs - no
sharefs - /etc/dfs/sharetab sharefs - no
/usr/lib/libc/libc_hwcap2.so.1 - /lib/libc.so.1 lofs - no
fd - /dev/fd fd - no rw
rpool/ROOT/zfsBE/var - /var zfs - no
rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime
swap - /tmp tmpfs - no xattr
rpool/VARSHARE - /var/share zfs - no
rw,devices,nosetuid,nonbmand,noexec,rstchown,noxattr,atime
rpool/export - /export zfs - no rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime
rpool/export/home - /export/home zfs - no
rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime
rpool/export/home/jack - /export/home/jack zfs - no
rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime
rpool - /rpool zfs - no rw,devices,setuid,nonbmand,exec,rstchown,xattr,atime
/dev/dsk/c2t1d0s0 - /mnt ufs - no rw,intr,largefiles,logging,xattr,onerror=panic

Unmounting a File System

You can use the unmount command to unmount a file system. For more information about
unmounting a ZFS file system, see “Unmounting ZFS File Systems” in Managing ZFS File
Systems in Oracle Solaris 11.4.

You can either unmount a specific file system or unmount a specific device that contains a file
system. For more information, see the umount(8) man page.

Ensure that the file system is not busy. A file system is considered busy, if a user is accessing
a directory in the file system, or if a program has an open file in that file system, or if the file
system is being shared.

Perform the following operations to make a file system available for unmounting:

■ Change to a directory in a different file system.

■ Stop the processes from using files on the file system.
■ Unshare the file system. For more information about unsharing a file system, see the
unshare(8) man page.

To verify that you have unmounted a file system, use the mount command.

$ mount | grep unmounted-file-system

Chapter 2 • Managing File Systems 33

How to Stop All Processes That Are Accessing a File System

How to Stop All Processes That Are Accessing a

File System
1. Become an administrator.
For more information, see “Using Your Assigned Administrative Rights” in Securing Users and
Processes in Oracle Solaris 11.4.

2. List all the processes that are accessing the file system to know the processes
that you want to stop.

# fuser -c [ -u ] /mount-point

-c Reports on files that are mount points for file systems and any files
within those mounted file systems

-u Displays the user login name for each process ID

/mount-point Specifies the name of the file system for which you want to stop
processes

3. Stop all processes that are accessing the file system.

# fuser -c -k /mount-point

A SIGKILL is sent to each process that is using the file system.

Note - You must not stop the processes of a user without first warning the user.

4. Verify that no process is accessing the file system.

# fuser -c /mount-point

Example 11 Stopping All Processes That Are Accessing a File System

This example shows how to stop process 4006c that is using the /export/home file system.

# fuser -c /export/home
/export/home: 4006c
# fuser -c -k /export/home
/export/home: 4006c
# fuser -c /export/home
/export/home:

34 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Checking UFS File System Consistencies

Checking UFS File System Consistencies

You might need to interactively check a file system when you are unable to mount it or when
the file system develops inconsistencies.

When a file system in use develops inconsistencies, error messages are displayed on the console
window or on the system message file, or the system might crash. For example, the system
messages file, /var/adm/messages, might include messages such as:

Sep 5 13:42:40 hostname ufs: [ID 879645 kern.notice] NOTICE: /: unexpected

free inode 630916, run fsck(1M)

hostname is the system which is reporting the error.

Consider the following aspects when running the fsck command to check UFS file systems:

■ A file system must be inactive in order to be checked. File system changes waiting to be
flushed to disk or file system changes that occur during fsck checking process can be
interpreted as file system corruption. These issues might not be a reliable indication of a
problem.
■ A file system must be inactive in order to be repaired. File system changes waiting to be
flushed to disk or file system changes that occur during the fsck repairing process might
cause the file system corruption. This might also cause the system to crash.
■ Unmount a file system before you use fsck on that file system. Unmounting the file system
ensures the data structures in it are consistent.

If a UFS file system runs out of space, you can see the following message in the /var/adm/
messages file.

file system full

For more information about troubleshooting file system issues, see Chapter 3, “Troubleshooting
File System Problems” in Troubleshooting System Administration Issues in Oracle Solaris 11.4.

How to Check a UFS File System

There is no need to rerun the fsck command if you see the following message:

***** FILE SYSTEM WAS MODIFIED *****

This message is an informational message about the actions of fsck. You can run fsck after the
message is displayed.

Chapter 2 • Managing File Systems 35

How to Check a UFS File System

1. Become an administrator.
For more information, see “Using Your Assigned Administrative Rights” in Securing Users and
Processes in Oracle Solaris 11.4.

2. Unmount the local file system to ensure that there is no activity on the file
system.
Specify the mount point directory or /dev/dsk/device-name as arguments to the fsck
command. If there are any inconsistencies, error messages are displayed.

# umount /export/home
# fsck /dev/rdsk/c0t0d0s7
** /dev/dsk/c0t0d0s7
** Last Mounted on /export/home
.
.

3. Correct any reported fsck errors.

For information about how to respond to the error message prompts while you check one or
more UFS file systems, see “Resolving UFS File System Inconsistencies” on page 38.

4. Mount the repaired file system to determine if there are any files in the lost+found
directory.
The fsck command moves individual files to the lost+found directory and these files are
renamed with their inode numbers.

5. Rename and move any files in the lost+found directory.

Rename the files and move them to their original location. Use the grep command to match
phrases within individual files and the file command to identify the type of file.
Remove all the unidentifiable files or directories left in the lost+found directory so that the
directory does not fill.
If fsck cannot repair all of the problems, see “Fixing a UFS File System That the fsck
Command Cannot Repair” on page 59.

Example 12 Interactive Checking of UFS File Systems

The following example shows how to check the /dev/rdsk/c0t0d0s6 file system and correct
the incorrect block count. This example assumes that the file system is unmounted.

# fsck /dev/rdsk/c0t0d0s6
** Phase 1 - Check Block and Sizes

36 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

INCORRECT BLOCK COUNT I=2529 (6 should be 2)

CORRECT? y

** Phase 2 - Check Pathnames

** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Cylinder Groups
929 files, 8928 used, 2851 free (75 frags, 347 blocks, 0.6%
fragmentation)

***** FILE SYSTEM WAS MODIFIED ****

Preening UFS File Systems

The fsck -o p command checks a UFS file system and automatically fix the problems that
normally result from an unexpected system shutdown. This command exits immediately if it
encounters a problem that requires operator intervention. This command also permits parallel
checking of the file systems.

You can run the fsck -o p command to preen the file systems after an unclean shutdown. In
this mode, the fsck command does not look at the clean flag and performs a full check. These
actions are a subset of the actions performed by the interactive fsck command.

How to Preen a UFS File System

1. Become an administrator.
For more information, see “Using Your Assigned Administrative Rights” in Securing Users and
Processes in Oracle Solaris 11.4.

2. Unmount the UFS file system.

# umount /mount-point

3. Check the UFS file system using the preen (-p) option.
# fsck -o p /dev/rdsk/device-name

You can preen individual file systems by using /mount-point or /dev/rdsk/device-name as

arguments to the fsck command.

Example 13 Preening a UFS File System

The following example shows how to preen the /export/home file system.

Chapter 2 • Managing File Systems 37

How to Preen a UFS File System

# fsck -o p /export/home

Resolving UFS File System Inconsistencies

The fsck command is run non-interactively to preen the file systems after an abrupt system
halt, where the latest file system changes are not written to disk. Preening automatically fixes
any basic file system inconsistencies and does not repair serious errors. While preening a file
system, the fsck command fixes the inconsistencies it expects from such an abrupt halt. For
serious conditions, the command reports the error and terminates.

When you run the fsck command interactively, it reports each inconsistency found and fixes
non-fatal errors. However, the command reports the inconsistency and prompts you to choose
a response for serious errors. When you run the fsck command using the -y or -n option, your
response is predefined as yes or no to the default response suggested by the fsck command for
each error condition.

A few corrective actions might result in some loss of data. You can determine the amount and
severity of data loss from the fsck diagnostic output.

The fsck command is a multipass file system check program. Each pass invokes a different
phase of the fsck command with different set of messages. After initialization, the fsck
command performs successive passes over each file system, checking blocks and sizes, path
names, connectivity, reference counts, and the map of free blocks (possibly rebuilding it). It also
performs cleanup.

The passes (phases) performed by the UFS version of the fsck command are as follows:
■ Initialization
■ Phase 1 – Check blocks and sizes
■ Phase 2a – Check duplicated names
■ Phase 2b – Check pathnames
■ Phase 3 – Check connectivity
■ Phase 3b – Verify shadows and access control lists (ACLs)
■ Phase 4 – Check reference counts
■ Phase 5 – Check cylinder groups

The following sections describe the error conditions that might be detected in each phase, the
messages and prompts that result, and the possible responses you can provide.

Messages that might appear in more than one phase are described in “General fsck Error
Messages” on page 39. Other messages are organized alphabetically by the phases in which
they occur.

38 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

The abbreviations that appear in the fsck command error messages are listed as follows:

BLK Block number

DUP Duplicate block number

DIR Directory name

CG Cylinder group

MTIME Time when the file was last modified

UNREF Unreferenced

General fsck Error Messages

The error messages in this section might be displayed in any phase after initialization. Although
they offer the option to continue, it is a best practice to consider them as fatal. They reflect a
serious system failure and must be handled immediately. When confronted with such messages,
terminate the program by entering n (no). If you are unable to determine the cause for the
problem, contact your local service provider.

CANNOT SEEK: BLK block-number (CONTINUE)

Cause: A request to move to the specified block number, in the file system failed. This
message indicates a serious hardware failure. If you want to continue the file system check,
fsck retries to move the specified block number and displays a list of sector numbers that
cannot be moved. If the block is a part of the virtual memory buffer cache, fsck terminates
with a fatal I/O error message.

Solution: If
the disk is experiencing hardware problems, the problem will persist. Run the
fsck command again to recheck the file system. If the recheck fails, contact your local
service provider.

CANNOT READ: DISK BLOCK block-number: I/O ERROR

CONTINUE?

Cause: A request to read a specified block number in the file system failed. The message
indicates a serious problem such as hardware failure. If you want to continue the file system
check, fsck retries to read the specified block number and displays a list of sector numbers
that cannot be read. If the block is a part of the virtual memory buffer cache, fsck terminates
with a fatal I/O error message. If fsck tries to write back one of the blocks on which the read
failed, it displays the following message:

Chapter 2 • Managing File Systems 39

How to Preen a UFS File System

WRITING ZERO'ED BLOCK sector-numbers TO DISK

Solution: If
the disk is experiencing hardware problems, the problem will persist. Run fsck
again to check the file system. If the check fails, contact your local service provider.

CANNOT WRITE: BLK block-number (CONTINUE)

Cause: A request to write a specified block number in the file system failed. If you continue
the file system check, fsck retries to write the specified block number and displays a list
of sector numbers that cannot be written. If the block is a part of the virtual memory buffer
cache, fsck terminates with a fatal I/O error message.

Solution: The disk might be write-protected. Check the write-protect lock on the drive. If
the disk has hardware problems, the problem will persist. Run fsck again to check the file
system. If write-protect is not the problem or if the check fails, contact your local service
provider.

Initialization Phase fsck Messages

In the initialization phase, command-line syntax is checked. Before the file system check can be
performed, the fsck command sets up tables and opens files.

The messages in this section relate to error conditions resulting from command-line options,
memory requests, opening of files, status of files, file system size checks, and the creation of the
scratch file. All such initialization errors terminate fsck when it is preening the file system.

Can't roll the log for device-name.

DISCARDING THE LOG MAY DISCARD PENDING TRANSACTIONS.

DISCARD THE LOG AND CONTINUE?

Cause: The fsck command was unable to flush the transaction log of a logging UFS file
system prior to checking the file system for errors.

Solution: If
you answer yes and continue, the file system operations that are in the log, but
have not been applied to the file system, are lost. In this case, fsck runs the same checks it
always runs and asks the following question in phase 5:

FREE BLK COUNT(S) WRONG IN SUPERBLK (SALVAGE)

Answering yes reclaims the blocks that were used for the log. The next time the file system
is mounted with logging enabled, the log is recreated. Answering no preserves the log and
exits, but the file system is not mountable.

40 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

bad inode number inode-number to ginode

Cause: An internal error occurred because of a nonexistent inode. This error exits fsck.

Solution: Contact your local service provider.

cannot alloc size-of-block map bytes for blockmap

cannot alloc size-of-free map bytes for freemap
cannot alloc size-of-state map bytes for statemap
cannot alloc size-of-lncntp bytes for lncntp

Cause: Request for memory for its internal tables failed. This failure terminates fsck. The
error message indicates a serious system failure that must be handled immediately. This
condition might occur if other processes are using a large amount of system resources.

Solution: Killing other processes might solve the problem. If not, contact your local service
provider.

Can't open checklist file: filename

Cause: The file system checklist file, (/etc/vfstab) cannot be opened to read. This failure
terminates fsck.

Solution: Check if the file exists and if the access modes of the file permits read access.

Can't open filename

Cause: The fsck command cannot open the file system, filename. When fsck is running in
an interactive mode, it ignores this file system and continues to check the next file system.

Solution: Check to see if read and write access to the raw device file for the file system is
permitted.

Can't stat root

Cause: The fsck command request for statistics about the root directory failed. This failure
terminates fsck.

Solution: This message indicates a serious system failure. Contact your local service
provider.

Can't stat filename

Can't make sense out of name filename

Chapter 2 • Managing File Systems 41

How to Preen a UFS File System

Cause: The fsck command request for statistics about the file system, filename failed. When
fsck is running in an interactive mode, it ignores this file system and continues to check the
next given file system.

Solution: Check if the file system exists and check its access modes.

filename: (NO WRITE)

Cause: Eitherthe -n option is specified or fsck cannot open the file system, filename to
write. When fsck is running in no-write mode, all diagnostic messages are displayed, but
fsck does not attempt to fix any errors.

Solution: If
the -n option is not specified, check the type of the file specified. It might be the
name of a regular file.

IMPOSSIBLE MINFREE=percent IN SUPERBLOCK (SET TO DEFAULT)

Cause: The superblock minimum space percentage is greater than 99 percent or less than 0
percent.

Solution: Toset the minfree parameter to the default 10 percent, type y at the default prompt.
To ignore the error condition, type n at the default prompt.

filename: BAD SUPER BLOCK: message

USE AN ALTERNATE SUPER-BLOCK TO SUPPLY NEEDED INFORMATION;
e.g., fsck[-f ufs] -o b=# [special ...]
where # is the alternate superblock. See fsck_ufs(1M)

Cause: The superblock is corrupted.

Solution: One of the following messages might be displayed:

CPG OUT OF RANGE

FRAGS PER BLOCK OR FRAGSIZE WRONG

INODES PER GROUP OUT OF RANGE

INOPB NONSENSICAL RELATIVE TO BSIZE

MAGIC NUMBER WRONG

NCG OUT OF RANGE

42 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

NCYL IS INCONSISTENT WITH NCG*CPG

NUMBER OF DATA BLOCKS OUT OF RANGE

NUMBER OF DIRECTORIES OUT OF RANGE

ROTATIONAL POSITION TABLE SIZE OUT OF RANGE

SIZE OF CYLINDER GROUP SUMMARY AREA WRONG

SIZE TOO LARGE BAD VALUES IN SUPERBLOCK

Try to rerun fsck with an alternate superblock. Specifying block 32 is a good choice. You
can locate an alternate copy of the superblock by running the newfs -N command on the
slice. If you do not specify the -N option, the newfs command overwrites the existing file
system.

UNDEFINED OPTIMIZATION IN SUPERBLOCK (SET TO DEFAULT)

Cause: The superblock optimization parameter (OPT_TIME or OPT_SPACE) is not defined.

Solution: To
minimize performance time of operations on the file system, type y at the SET
TO DEFAULT prompt. To ignore this error condition, type n.

Phase 1 – Check Blocks and Sizes Messages

This phase checks the inode list. It reports error in the following conditions:
■ Checking inode types
■ Setting up the zero-link-count table
■ Examining inode block numbers for bad or duplicate blocks
■ Checking inode size
■ Checking inode format

All errors in this phase except INCORRECT BLOCK COUNT, PARTIALLY TRUNCATED INODE,
PARTIALLY ALLOCATED INODE, and UNKNOWN FILE TYPE terminate fsck when it is preening a
file system.

The following messages might occur in phase 1 of fsck check.

block-number BAD I=inode-number

Cause: The inode, inode-number contains a block number, block-number, with a number
lower than the number of the first data block in the file system or greater than the number

Chapter 2 • Managing File Systems 43

How to Preen a UFS File System

of the last block in the file system. This error condition might generate the EXCESSIVE BAD
BLKS error message in phase 1, if inode has too many block numbers outside the file system
range. This generates BAD/DUP error message in phases 2 and 4.

BAD MODE: MAKE IT A FILE?

Cause: The status of a given inode is set to 1s, indicating file system damage. This message
does not indicate physical disk damage, unless it is displayed repeatedly after running the
fsck -y command.

Solution: Type y to reinitialize the inode to a reasonable value.

BAD STATE state-number TO BLKERR

Cause: An internal error has scrambled the fsck state map so that it shows the impossible
value of state-number. This error exits fsck immediately.

Solution: Contact your local service provider.

fragment-number DUP I=inode-number

Cause: The inode inode-number, contains a block number, fragment-number, which is

already claimed by the same inode or another inode. If inode has too many block numbers
claimed by the same inode or another inode, this error condition might generate EXCESSIVE
DUP BLKS error message in phase 1. This error condition invokes phase 1B and generates the
BAD/DUP error messages in phases 2 and 4.

DUP TABLE OVERFLOW (CONTINUE)

Cause: The fsck command is unable to allocate memory to track duplicate fragments. If the
-o p option is specified, the program terminates.

Solution: To continue the program, type y at the CONTINUE prompt. When this error occurs,
a complete check of the file system is not possible. If another duplicate fragment is found,
this error condition repeats. Increase the amount of virtual memory available by killing some
processes or increase swap space to run fsck again to check the file system. To terminate the
program, type n.

EXCESSIVE BAD FRAGMENTS I=inode-number (CONTINUE)

Cause: Too many (more than 10) fragments indicate an invalid disk address. If the -o p
option is specified, the program terminates.

44 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

Solution: Tocontinue the program, type y at the CONTINUE prompt. When this error occurs, a
complete check of the file system is not possible. Run fsck again to check the file system.
To terminate the program, type n.

EXCESSIVE DUP BLKS I=inode-number (CONTINUE)

Cause: Too many (more than 10) fragments are claimed by the same inode, or another inode,
or by a free-list. If the -o p option is specified, the program terminates.

Solution: Tocontinue the program, type y at the CONTINUE prompt. When this error occurs, a
complete check of the file system is not possible. Run fsck again to check the file system.
To terminate the program, type n.

INCORRECT BLOCK COUNT I=inode-number (number-of-BAD-DUP-or-missing-blocks

should be number-of-blocks-in-filesystem) (CORRECT)

Cause: The disk block count for inode, inode-number is incorrect. When preening, fsck
corrects the count.

Solution: To correct the disk block count of inode, inode-number by number of blocks in
the file system, type y at the CORRECT prompt. Typing y at the CORRECT prompt replaces the
block count of inode inode-number by number of blocks in the file system. To terminate the
program, type n.

LINK COUNT TABLE OVERFLOW (CONTINUE)

Cause: There is no more room in the internal table for fsck containing allocated inodes
with a link count of zero. If the -o p option is specified, the program exits and fsck must be
completed manually.

Solution: To continue the program, type y at the CONTINUE prompt. If another allocated
inode with a zero-link count is found, this error condition repeats. When this error occurs, a
complete check of the file system is not possible. Run fsck again to recheck the file system.
Increase the virtual memory available by killing some processes or increasing swap space,
and then run fsck again. To terminate the program, type n.

PARTIALLY ALLOCATED INODE I=inode-number (CLEAR)

Cause: The inode, inode-number is not allocated or unallocated. If the -o p option is

specified, the inode is cleared.

Solution: Todeallocate the inode, inode-number by zeroing out its contents, type y. Typing y
generates the UNALLOCATED error condition in phase 2 for each directory entry pointing to this

Chapter 2 • Managing File Systems 45

How to Preen a UFS File System

inode. To ignore the error condition, type n. A no response is appropriate only if you intend
to take other measures to fix the problem.

PARTIALLY TRUNCATED INODE I=inode-number (SALVAGE)

Cause: The fsck command has found inode, inode-number whose size is shorter than
the number of fragments allocated to it. This condition occurs only if the system crashes
while truncating a file. When preening the file system, fsck completes the truncation to the
specified size.

Solution: Tocomplete the truncation to the size specified in the inode, type y at the SALVAGE
prompt. To ignore this error condition, type n.

UNKNOWN FILE TYPE I=inode-number (CLEAR)

Cause: The mode word of the inode, inode-number shows that the inode is not a pipe,
character device, block device, regular file, symbolic link, FIFO file, or a directory inode. If
the -o p option is specified, the inode is cleared.

Solution: To deallocate the inode, inode-number by zeroing its contents, results in the
UNALLOCATED error condition in phase 2 for each directory entry pointing to this inode, type y
at the CLEAR prompt. To ignore this error condition, type n.

Phase 1B – Rescan for More DUPS Messages

When a duplicate block is found in a file system, the following message is displayed.

block-number DUP I=inode-number

Cause: The inode, inode-number contains a block number, block-number that is already
claimed by the same inode or another inode. This error condition generates the BAD/DUP error
message in phase 2. Inodes that have overlapping blocks might be determined by examining
this error condition and the DUP error condition in phase 1.

Solution: When a duplicate block is found, the file system is rescanned to find the inode that
previously claimed that block.

Phase 2 – Check Path Names Messages

The phase 2 of the fsck check removes directory entries pointing to bad inodes found in phases
1 and 1B. It reports an error in the following conditions:

46 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

■ Incorrect root inode mode and status

■ Directory inode pointers out of range
■ Directory entries pointing to bad inodes
■ Directory integrity checks

When the file system is being preened using the -o p option, all errors in this phase terminate
fsck, except those related to directories not being a multiple of the block size, duplicate and bad
blocks, inodes out of range, and extraneous hard links.

The following messages might occur in phase 2 of fsck check.

BAD INODE state-number TO DESCEND

Cause: An fsck internal error has passed an invalid state, state-number to the routine that
descends the file system directory structure. This error exits fsck.

Solution: If this error message is displayed, contact your local service provider.

BAD INODE NUMBER FOR '.' I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (FIX)

Cause: A directory inode-number has been found whose inode number for the first entry in a
directory (.) does not match inode-number.

Solution: Tochange the inode number for . and match the inode-number, type y at the FIX
prompt. If you do not want to change the inode number for ., type n.

BAD INODE NUMBER FOR '..' I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (FIX)

Cause: A directory inode-number has been found whose inode number for the second entry
in a directory (..) does not match the parent of inode-number.

Solution: To set the inode number for .. to match the parent directory of inode-number, type
y at the FIX prompt. The .. in the root inode points to itself. If you do not want to change the
inode number for .., type n.

BAD RETURN STATE state-number FROM DESCEND

Cause: An fsck internal error has returned an impossible state, state-number from the routine
that descends the file system directory structure. This error exits fsck.

Solution: If this message is displayed, contact your local service provider.

Chapter 2 • Managing File Systems 47

How to Preen a UFS File System

BAD STATE state-number FOR ROOT INODE

Cause: An internal error has assigned an impossible state, state-number to the root inode.
This error exits fsck.

Solution: If this error message is displayed, contact your local service provider.

BAD STATE state-number FOR INODE=inode-number

Cause: Aninternal error has assigned an impossible state, state-number to inode, inode-
number. This error exits fsck.

Solution: If this error message is displayed, contact your local service provider.

DIRECTORY TOO SHORT I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (FIX)

Cause: A directory filename has been found whose size, file-size is less than the minimum
directory size. The owner UID, mode file-mode, size file-size, modify time modification-
time, directory name, and filename are displayed.

Solution: Toincrease the size of the directory to the minimum directory size, type y at the FIX
prompt. To ignore this directory, type n.

DIRECTORY filename: LENGTH file-size NOT MULTIPLE OF disk-block-size (ADJUST)

Cause: A directory filename has been found with size, file-size that is not a multiple of the
directory block size, disk-block-size.

Solution: To round up the length to the appropriate disk block size, type y. When preening the
file system using the -o p option, fsck only displays a warning and adjusts the directory. To
ignore this condition, type n.

DIRECTORY CORRUPTED I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (SALVAGE)

Cause: A directory with an inconsistent internal state has been found.

Solution: To throw away all entries up to the next directory boundary (usually a 512-byte
boundary), type y at the SALVAGE prompt. This drastic action can throw away up to 42
entries. You must perform this action only after other recovery efforts have failed. To skip to
the next directory boundary and resume reading, but not modify the directory, type n.

DUP/BAD I=inode-number OWNER=O MODE=M SIZE=file-size

MTIME=modification-time TYPE=filename (REMOVE)

48 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

Cause: Phase 1 or phase 1B has found duplicate fragments or bad fragments associated with
directory or file entry filename, inode inode-number. The owner UID, mode file-mode,
size file-size, modification time modification-time, and directory or file name filename are
displayed. If the -o p option is specified, the duplicate or bad fragments are removed.

Solution: To remove the directory or file entry filename, type y at the REMOVE prompt. To
ignore this error condition, type n.

DUPS/BAD IN ROOT INODE (REALLOCATE)

Cause: Phase 1 or phase 1B has found duplicate fragments or bad fragments in the root
inode (usually inode-number 2 of the file system).

Solution: To clear the existing contents of the root inode and reallocate it, type y at the
REALLOCATE prompt. The files and directories usually found in the root inode are recovered
in phase 3 and moved to the lost+found directory. If the attempt to allocate the root
fails, fsck exits with an error message, CANNOT ALLOCATE ROOT INODE. Type n to get the
CONTINUE prompt. Type y to respond to the CONTINUE prompt, and ignore the DUPS or BAD
error condition in the root inode and to continue the file system check. If the root inode is
not correct, this might generate many other error messages. Type n to terminate the program.

EXTRA '.' ENTRY I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (FIX)

Cause: More than one entry for . in a inode-number directory has been found.

Solution: Toremove the extra entry for ., type y at the FIX prompt. To leave the directory
unchanged, type n.

EXTRA '..' ENTRY I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (FIX)

Cause: A directory inode-number has been found that has more than one entry for .. in the
parent directory.

Solution: Toremove the extra entry for .. in the parent directory, type y at the FIX prompt.
To leave the directory unchanged, type n.

hard-link-number IS AN EXTRANEOUS HARD LINK TO A DIRECTORY file-name (REMOVE)

Cause: The fsck command has found an extraneous hard link, hard-link-number to a
directory file-name. When preening using the -o p option, fsck ignores the extraneous hard
links.

Chapter 2 • Managing File Systems 49

How to Preen a UFS File System

Solution: To delete the extraneous entry hard-link-number, type y at the REMOVE prompt. To
ignore the error condition, type n.

inode-number OUT OF RANGE I=inode-number NAME=filename (REMOVE)

Cause: A directory entry filename has an inode number, inode-number that is greater
than the end of the inode list. If the -p option of preen is specified, the inode is removed
automatically.

Solution: To delete the directory entry filename, type y at the REMOVE prompt. To ignore the
error condition, type n.

MISSING '.' I=inode-number OWNER=UID MODE=file-mode SIZE=file-sizee

MTIME=modification-time DIR=filename (FIX)

Cause: A directory, inode-number has been found whose first entry (.) is unallocated.

Solution: Tobuild an entry for . with inode number equal to inode-number, type y at the FIX
prompt. To leave the directory unchanged, type n.

MISSING '.' I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time DIR=filename CANNOT FIX, FIRST ENTRY IN
DIRECTORY CONTAINS filename

Cause: A
directory, inode-number has been found whose first entry is filename. The fsck
command cannot resolve this problem.

Solution: If this error message is displayed, contact your local service provider.

MISSING '.' I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time DIR=filename CANNOT FIX, INSUFFICIENT
SPACE TO ADD '.'

Cause: A directory inode-number has been found whose first entry is not pointing to the
same directory (.). The fsck command cannot resolve the problem.

Solution: If this error message is displayed, contact your local service provider.

MISSING '..' I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time DIR=filename (FIX)

Cause: A directory, inode-number has been found whose second entry (..) is unallocated.

Solution: To build an entry for .. with inode number equal to the parent of inode-number,
type y at the FIX prompt. The .. in the root inode points to itself. To leave the directory
unchanged, type n.

50 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

MISSING '..' I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time DIR=filename CANNOT FIX, SECOND ENTRY IN
DIRECTORY CONTAINS filename

Cause: A
directory, inode-number has been found whose second entry is filename. The fsck
command cannot resolve this problem.

Solution: If this error message is displayed, contact your local service provider.

MISSING '..' I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time DIR=filename CANNOT FIX, INSUFFICIENT SPACE
TO ADD '..'

Cause: A directory, inode-number has been found whose second entry is not .. in the parent
directory. The fsck command cannot resolve this problem.

Solution: If this error message is displayed, contact your local service provider.

NAME TOO LONG filename

Cause: An excessively long path name has been found, which usually indicates loops in the
file system name space. This error can occur if a privileged user has made circular links to
directories.

Solution: Remove the circular links using the rm command.

ROOT INODE UNALLOCATED (ALLOCATE)

Cause: The root inode (mostly inode number 2) has no allocate-mode bits.

Solution: To allocate inode 2 as the root inode, type y at the ALLOCATE prompt. The files
and directories usually found in the root inode are recovered in phase 3 and moved to the
lost+found directory. If the attempt to allocate the root inode fails, fsck displays CANNOT
ALLOCATE ROOT INODE message and exits. To terminate the program, type n.

ROOT INODE NOT DIRECTORY (REALLOCATE)

Cause: The root inode (mostly inode number 2) of the file system is not a directory inode.

Solution: To clear the existing contents of the root inode and reallocate it, type y at the
REALLOCATE prompt. The files and directories usually found in the root inode are recovered
in phase 3 and moved to the lost+found directory. If the attempt to allocate the root inode
fails, fsck displays the message, CANNOT ALLOCATE ROOT INODE and exits. To have fsck
prompt with FIX to try to repair the inode, type n.

Chapter 2 • Managing File Systems 51

How to Preen a UFS File System

UNALLOCATED I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time type=filename(REMOVE)

Cause: A directory or file entry filename points to an unallocated inode, inode-number. The
owner UID, mode file-mode, size file-size, modify time modification-time, and file name
filename are displayed.

Solution: To delete the directory entry filename, type y at the REMOVE prompt. To ignore the
error condition, type n.

ZERO LENGTH DIRECTORY I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time DIR=filename (REMOVE)

Cause: A directory entry filename has a size, file-size that is zero. The owner UID, mode
file-mode, size file-size, modify time modification-time, and directory name filename are
displayed.

Solution: To remove the directory entry filename, type y at the REMOVE prompt. This results in
the BAD or DUP error message in phase 4. To ignore the error condition, type n.

Phase 3 – Check Connectivity Messages

This phase checks the directories examined in phase 2 and reports error conditions resulting
from unreferenced directories, and missing or full lost+found directory.

The following messages might occur in phase 3 fsck check.

BAD INODE state-number TO DESCEND

Cause: An internal error has caused an impossible state, state-number to be passed to the
routine that descends the file system directory structure. The error terminates fsck.

Solution: If this occurs, contact your local service provider.

DIR I=inode-number1 CONNECTED. PARENT WAS I=inode-number2

Cause: This is an advisory message indicating a directory inode, inode-number1 was

successfully connected to the lost+found directory. The parent inode, inode-number2 of
the directory inode, inode-number1 is replaced by the inode number of the lost+found
directory.

DIRECTORY filename LENGTH file-size NOT MULTIPLE OF disk-block-size (ADJUST)

52 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

Cause: A directory filename has been found with size, file-size that is not a multiple of the
disk-block-size. This condition can recur in phase 3 if it is not corrected in phase 2.

Solution: To round up the length to the appropriate disk block size, type y at the ADJUST
prompt. When preening, fsck displays a warning and adjusts the directory. To ignore this
error condition, type n.

lost+found IS NOT A DIRECTORY (REALLOCATE)

Cause: The entry for lost+found is not a directory.

Solution: To allocate a directory inode and change the lost+found directory to reference it,
type y at the REALLOCATE prompt. The previous inode reference by the lost+found directory
is not cleared and it reclaims as an unreferenced inode or has its link count adjusted later
in this phase. Inability to create a lost+found directory displays the message, SORRY.
CANNOT CREATE lost+found DIRECTORY and aborts the attempt to link the lost inode, which
generates the UNREF error message in phase 4. To abort the attempt to link the lost inode,
which generates the UNREF error message in phase 4, type n.

NO lost+found DIRECTORY (CREATE)

Cause: There is no lost+found directory in the root directory of the file system. When
preening, fsck tries to create a lost+found directory.

Solution: Tocreate a lost+found directory in the root of the file system, type y at the
CREATE prompt. This might lead to the message NO SPACE LEFT IN / (EXPAND). If the lost
+found directory cannot be created, fsck displays the message, SORRY. CANNOT CREATE
lost+found DIRECTORY and aborts the attempt to link the lost inode. This in turn generates
the UNREF error message later in phase 4. To abort the attempt to link up the lost inode, type n
at the CREATE prompt.

NO SPACE LEFT IN /lost+found (EXPAND)

Cause: Another entry cannot be added to the lost+found directory in the root directory of
the file system because there is no available space. When preening, fsck expands the lost
+found directory.

Solution: Toexpand the lost+found directory and to make room for the new entry, type y
at the EXPAND prompt. If the attempted expansion fails, fsck displays a message, SORRY.
NO SPACE IN lost+found DIRECTORY and aborts the request to link a file to the lost
+found directory. This error generates the UNREF error message later in phase 4. Delete any

Chapter 2 • Managing File Systems 53

How to Preen a UFS File System

unnecessary entries in the lost+found directory. This error terminates fsck when preening is
in effect. To abort the attempt to link the lost inode, type n at the EXPAND prompt.

UNREF DIR I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time (RECONNECT)

Cause: The directory inode, inode-number was not connected to a directory entry when the
file system was traversed. The owner UID, mode file-mode, size file-size, and modification
time modification-time of directory inode inode-number are displayed. When preening, fsck
reconnects the non-empty directory inode if the directory size is non-zero. Otherwise, fsck
clears the directory inode.

Solution: To reconnect the directory inode, inode-number into the lost+found directory,
type y at the RECONNECT prompt. If the directory is successfully reconnected, a CONNECTED
message is displayed. Otherwise, one of the lost+found error messages is displayed. To
ignore this error condition, type n at the RECONNECT prompt. This error causes the UNREF error
condition in phase 4.

Phase 4 – Check Reference Counts Messages

This phase checks the link count information obtained in phases 2 and 3. It reports error
conditions resulting from:

■ Unreferenced files
■ A missing or full lost+found directory
■ Incorrect link counts for files, directories, symbolic links, or special files
■ Unreferenced files, symbolic links, and directories
■ Bad or duplicate blocks in files and directories
■ Incorrect total free-inode counts

All errors in this phase (except running out of space in the lost+found directory) are
correctable when the file system is being preened.

The following messages might occur in phase 4.

BAD/DUP type I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time (CLEAR)

Cause: Phase 1 or phase 1B found duplicate blocks or bad blocks associated with file
or directory inode, inode-number. The owner UID, mode file-mode, size file-size, and
modification time modification-time of inode, inode-number are displayed.

54 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

Solution: Todeallocate inode, inode-number by zeroing its contents, type y at the CLEAR
prompt. To ignore this error condition, type n.

(CLEAR)

Cause: The inode mentioned in the UNREF error message preceding immediately cannot be
reconnected. This message is not displayed if the file system is being preened because of
lack of space to reconnect to files. This terminates fsck.

Solution: To deallocate the inode by zeroing out its contents, type y at the CLEAR prompt. To
ignore the preceding error condition, type n.

LINK COUNT type I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size
MTIME=modification-time COUNT link-count SHOULD BE
corrected-link-count (ADJUST)

Cause: The link count for a directory or a file inode with inode-number is link-count
but must be corrected-link-count. The owner UID, mode file-mode, size file-size, and
modification time modification-time of the inode, inode-number are displayed. If the -o p
option is specified, the link count is adjusted unless the number of references is increasing.
This condition does not occur unless there is a hardware failure. When the number of
references is increasing during preening, fsck displays the LINK COUNT INCREASING
message and exits.

Solution: To replace the link count of directory or file inode inode-number with corrected-
link-count, type y at the ADJUST prompt. To ignore this error condition, type n.

lost+found IS NOT A DIRECTORY (REALLOCATE)

Cause: The entry for lost+found is not a directory.

Solution: To allocate a directory inode and change the lost+found directory to reference
it, type y at the REALLOCATE prompt. The previous inode referenced by the lost+found
directory is not cleared. It is either reclaimed as an unreferenced inode or has its link count
adjusted later in this phase. Inability to create a lost+found directory displays the message,
SORRY. CANNOT CREATE lost+found DIRECTORY and aborts the attempt to link up the lost
inode. This error generates the UNREF error message later in phase 4. To abort the attempt to
link up the lost inode, type n at the REALLOCATE prompt.

NO lost+found DIRECTORY (CREATE)

Cause: There is no lost+found directory in the root directory of the file system. When
preening, fsck tries to create a lost+found directory.

Chapter 2 • Managing File Systems 55

How to Preen a UFS File System

Solution: To create a lost+found directory in the root of the file system, type y at the CREATE
prompt. If the lost+found directory cannot be created, fsck displays the message, SORRY.
CANNOT CREATE lost+found DIRECTORY and aborts the attempt to link the lost inode. This
error in turn generates the UNREF error message later in phase 4. To abort the attempt to link
up the lost inode, type n at the CREATE prompt.

NO SPACE LEFT IN / lost+found (EXPAND)

Cause: There is no space to add another entry to the lost+found directory in the root
directory of the file system. When preening, fsck expands the lost+found directory.

Solution: To expand the lost+found directory in order to make room for the new entry, type
y at the EXPAND prompt. If the attempted expansion fails, fsck displays the message, SORRY.
NO SPACE IN lost+found DIRECTORY and aborts the request to link a file to the lost
+found directory. This error generates the UNREF error message later in phase 4. Delete any
unnecessary entries in the lost+found directory. This error terminates fsck when preening
using -o p option is in effect. To abort the attempt to link up the lost inode, type n at the
EXPAND prompt.

UNREF FILE I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time (RECONNECT)

Cause: File inode inode-number was not connected to a directory entry when the file system
was traversed. The owner UID, mode file-mode, size file-size, and modification time,
modification-time of inode, inode-number are displayed. When fsck is preening, the file is
cleared if either its size or its link count is zero; otherwise, it is reconnected.

Solution: To reconnect inode, inode-number to the file system in the lost+found directory,
type y. This error might generate the lost+found error message in phase 4 if there are
problems connecting inode, inode-number to the lost+found directory. To ignore this error
condition, type n. This error always invokes the CLEAR error condition in phase 4.

UNREF type I=inode-number OWNER=UID MODE=file-mode SIZE=file-size

MTIME=modification-time (CLEAR)

Cause: Inode, inode-number (whose type is directory or file) was not connected to a
directory entry when the file system was traversed. The owner UID, mode file-mode, size
file-size, and modification time modification-time of inode inode-number are displayed.
When fsck is preening, the file is cleared if either its size or its link count is zero; otherwise,
it is reconnected.

Solution: Todeallocate inode, inode-number by zeroing its contents, type y at the CLEAR
prompt. To ignore this error condition, type n at the CLEAR prompt.

56 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

ZERO LENGTH DIRECTORY I=inode-number OWNER=UID MODE=file-mode

SIZE=file-size MTIME=modification-time(CLEAR)

Cause: A directory entry filename has a size, file-size that is zero. The owner UID, mode file-
mode, size file-size, modification time modification-time, and directory name filename are
displayed.

Solution: Todeallocate the directory inode, inode-number by zeroing out its contents, type y.
To ignore the error condition, type n.

Phase 5 – Check Cylinder Groups Messages

This section contains phase 5 fsck messages.

This phase checks the free-fragment and used-inode maps. It reports error conditions resulting
from:
■ Allocated inodes missing from used-inode maps
■ Free fragments missing from free-fragment maps
■ Free inodes in the used-inode maps
■ Incorrect total free-fragment count
■ Incorrect total used inode count

The following messages might occur in phase 5.

FRAG BITMAP WRONG (CORRECTED)

Cause: A cylinder group (CG) fragment map is missing some free fragments. During
preening, fsck reconstructs the maps.

Solution: To reconstruct the free-fragment map, type y at the SALVAGE prompt. To ignore this
error condition, type n.

CG cg-number: BAD MAGIC NUMBER

Cause: The magic number of CG, cg-number is wrong. This error usually indicates that the
CG maps have been destroyed. When running interactively, the CG is marked as needing
reconstruction. fsck terminates if the file system is being preened.

Solution: If this occurs, contact your local service provider.

CORRECT GLOBAL SUMMARY (SALVAGE)

Chapter 2 • Managing File Systems 57

How to Preen a UFS File System

Cause: Thesummary information is incorrect. When preening, fsck recomputes the

summary information.

Solution: To reconstruct the summary information, type y at the SALVAGE prompt. To ignore
this error condition, type n at the SALVAGE prompt.

fsck Summary Messages

This section contains fsck summary messages in the current Oracle Solaris release. The
following messages are displayed in the cleanup phase.

Once a file system has been checked, a few summary messages are displayed.

number-of files, number-of-files

used, number-of-files free (number-of frags, number-of blocks,
percent fragmentation)

This message indicates that the file system checked contains the number of files using the
number of fragment-sized blocks, and that there are number of fragment-sized blocks free in
the file system. The numbers in parentheses break the free count down into number of free
fragments, number of free full-sized blocks, and the percent fragmentation.

***** FILE SYSTEM WAS MODIFIED *****

This message indicates that the file system was modified by fsck. There is no need to rerun
fsck if you see this message. This message is just informational about corrective actions
performed by fsck.

Cleanup Phase Messages

This section contains fsck cleanup phase messages in the Oracle Solaris 10 and later releases.
In this release, similar messages can be found in the fsck summary phase.

Once a file system has been checked, a few cleanup functions are performed. The cleanup phase
displays the following status messages.

number-of files, number-of-files

used, number-of-files free (number-of frags, number-of blocks,
percent fragmentation)

This message indicates that the file system checked contains number of files using number
of fragment-sized blocks, and that there are number of fragment-sized blocks free in the file

58 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Preen a UFS File System

system. The numbers in parentheses break the free count down into number of free fragments,
number-of free full-sized blocks, and the percent fragmentation.

***** FILE SYSTEM WAS MODIFIED *****

This message indicates that the file system was modified by fsck. If this file system is mounted
or is the current root (/) file system, reboot. If the file system is mounted, you might need to
unmount it and run fsck again; otherwise, the work done by fsck might be undone by the in-
core copies of tables.

filename FILE SYSTEM STATE SET TO OKAY

This message indicates that file system, filename was marked as stable. Use the fsck -m
command to determine if the file system needs checking.

filename FILE SYSTEM STATE NOT SET TO OKAY

This message indicates that file system, filename was not marked as stable. Use the fsck -m
command to determine if the file system needs checking.

Fixing a UFS File System That the fsck Command

Cannot Repair
The fsck command operates in several passes. For example, a problem corrected in later passes
can expose other problems that have occurred in earlier passes. Therefore, it is sometimes
necessary to run fsck until it no longer reports any problems. Doing so ensures that all errors
are found and repaired.

Pay attention to the information displayed by the fsck command. This information might help
you fix the problem. For example, the messages might point to a damaged directory. If you
delete the directory, you might find that the fsck command runs cleanly.

If the fsck command still cannot repair the file system, try to use the ff, clri, and ncheck
commands to understand and fix what is wrong. For information about these commands, see the
fsdb(8), ff(8), clri(8), and ncheck(8) man pages.

If you are unable to repair a file system but can mount it as read-only, try using the cp, tar, or
cpio commands to retrieve all or part of the data from the file system.

If hardware disk errors are causing the problem, you might need to reformat and repartition
the disk again before re-creating and restoring file systems. Check if the device cables and
connectors are functional before replacing the disk device. Hardware errors usually display the
same error again and again across different commands. The format command tries to work

Chapter 2 • Managing File Systems 59

How to Restore a Bad Superblock

around bad blocks on the disk. However, if the disk is too severely damaged, the problems
might persist, even after you reformat the disk. For information about the format command, see
the format(8) man page.

How to Restore a Bad Superblock

1. Become an administrator.
For more information, see “Using Your Assigned Administrative Rights” in Securing Users and
Processes in Oracle Solaris 11.4.

2. Change to a directory outside the damaged file system and unmount the file
system.

# unmount /mount-point

3. Display the superblock values using the newfs -N command.

# newfs -N /dev/rdsk/device-name

The command output displays the block numbers that were used for the superblock copies when
the newfs command created the file system, unless the file system was created with special
parameters. For information about creating a customized file system, “Customizing UFS File
System Parameters” on page 19.

Caution - Ensure to use the -N option. If you omit the -N option, you might destroy all the data
in the file system and replace it with an empty file system.

4. Provide an alternate superblock using the fsck command.

# fsck -F ufs -o b=block-number /dev/rdsk/device-name

The fsck command uses an alternate superblock to restore the primary superblock. You can
try 32 as an alternate block. You can use any of the alternate blocks shown by the newfs -N
command.

Example 14 Restoring a Bad Superblock

The following example shows how to restore the superblock copy 5264.

# newfs -N /dev/rdsk/c0t3d0s7
/dev/rdsk/c0t3d0s7: 163944 sectors in 506 cylinders of 9 tracks, 36 sectors
83.9MB in 32 cyl groups (16 c/g, 2.65MB/g, 1216 i/g)

60 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Restore a Bad Superblock

super-block backups (for fsck -b #) at:

32, 5264, 10496, 15728, 20960, 26192, 31424, 36656, 41888,
47120, 52352, 57584, 62816, 68048, 73280, 78512, 82976, 88208,
93440, 98672, 103904, 109136, 114368, 119600, 124832, 130064, 135296,
140528, 145760, 150992, 156224, 161456,
# fsck -F ufs -o b=5264 /dev/rdsk/c0t3d0s7
Alternate superblock location: 5264.
** /dev/rdsk/c0t3d0s7
** Last Mounted on
** Phase 1 - Check Blocks and Sizes
** Phase 2 - Check Pathnames
** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Check Cyl groups
36 files, 867 used, 75712 free (16 frags, 9462 blocks, 0.0% fragmentation)

***** FILE SYSTEM WAS MODIFIED *****

#

Chapter 2 • Managing File Systems 61

62 Managing File Systems in Oracle Solaris 11.4 • November 2020
 ♦ ♦ ♦
3
C H A P T E R 3

Managing Data and Free Space

When a file system runs out of free space, it leads to numerous system and application failures.
When the file system gets full, you might experience the following system issues:

■ Degradation in performance
■ Difficulty in spotting unused blocks by the kernel to store new data
■ System hangs due to long response time

The threshold when the effects on the system becomes visible varies from one file system to
another. As a best practice, always keep the file system space usage below 90%.

Oracle Solaris provides the following standard UNIX tools to monitor file system space usage:

df Displays the number of used blocks, number of free blocks, and

calculates total file system space usage in percentage.

du Displays the number of blocks a file or a set of files occupies.

File System Space Usage

The df tool without any options, prints the usage statistics for all mounted file systems. You can
use the -k and -h options for more descriptive output. The -k option changes block units to 1024
bytes (effectively kilobytes) and the -h option generates human readable output as the amount is
displayed in kilobytes, megabytes or gigabytes.

# df [-kh] [directory, dataset or device]

The df tool can use a directory, a ZFS dataset or a block device to filter the output. The
command then prints out the statistics just for the given file system.

Chapter 3 • Managing Data and Free Space 63

File System Space Usage

EXAMPLE 15 Displaying File System Space Usage

This example shows an output of the df tool. The output shows mount points, devices
containing the file system, number of free blocks (in units of 512 bytes), and the number of files
that can be stored.

# df
/ (rpool/ROOT/solaris):63807702 blocks 63807702 files
/var (rpool/ROOT/solaris/var):63807702 blocks 63807702 files
/devices (/devices ): 0 blocks 0 files
/dev (/dev ): 0 blocks 0 files
/system/contract (ctfs ): 0 blocks 2147483583 files
/proc (proc ): 0 blocks 29927 files
/etc/mnttab (mnttab ): 0 blocks 0 files
/system/volatile (swap ): 4176136 blocks 1130857 files
/tmp (swap ): 4176136 blocks 1130857 files
/system/object (objfs ): 0 blocks 2147483413 files
/etc/dfs/sharetab (sharefs ): 0 blocks 2147483646 files
/dev/fd (fd ): 0 blocks 0 files
/lib/libc.so.1 (/usr/lib/libc/libc_hwcap2.so.1):63807702 blocks 63807702 files
/var/share (rpool/VARSHARE ):63807702 blocks 63807702 files
/export (rpool/export ):63807702 blocks 63807702 files
/export/home (rpool/export/home ):63807702 blocks 63807702 files
/export/home/jack (rpool/export/home/jack):63807702 blocks 63807702 files
/rpool (rpool ):63807702 blocks 63807702 files
/system/zones (rpool/VARSHARE/zones):63807702 blocks 63807702 files
/var/share/kvol (rpool/VARSHARE/kvol):63807702 blocks 63807702 files
/var/share/sstore/repo(rpool/VARSHARE/sstore):63807702 blocks 63807702 files
/var/share/tmp (rpool/VARSHARE/tmp):63807702 blocks 63807702 files
/var/share/pkg (rpool/VARSHARE/pkg):63807702 blocks 63807702 files
/var/share/pkg/repositories(rpool/VARSHARE/pkg/repositories):63807702 blocks 63807702
files
/mnt (/dev/dsk/c2t2d0s0 ):107192852 blocks 8491364 files

EXAMPLE 16 File System Space Usage in Descriptive Format With Unit Suffixes

This example shows an output of the df command where numbers of blocks are translated. The
output also shows the file system usage in percentages.

# df -h
Filesystem Size Used Available Capacity Mounted on
rpool/ROOT/solaris 67G 3.4G 30G 10% /
rpool/ROOT/solaris/var
67G 574M 30G 2% /var
/devices 0K 0K 0K 0% /devices
/dev 0K 0K 0K 0% /dev
ctfs 0K 0K 0K 0% /system/contract

64 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Files Using the Most Space in a File System

proc 0K 0K 0K 0% /proc
mnttab 0K 0K 0K 0% /etc/mnttab
swap 2.0G 2.5M 2.0G 1% /system/volatile
swap 2.0G 60K 2.0G 1% /tmp
objfs 0K 0K 0K 0% /system/object
sharefs 0K 0K 0K 0% /etc/dfs/sharetab
fd 0K 0K 0K 0% /dev/fd
/usr/lib/libc/libc_hwcap2.so.1
34G 3.4G 30G 10% /lib/libc.so.1
rpool/VARSHARE 67G 77M 30G 1% /var/share
rpool/export 67G 36K 30G 1% /export
rpool/export/home 67G 33K 30G 1% /export/home
rpool/export/home/jack
67G 34K 30G 1% /export/home/jack
rpool 67G 3.3M 30G 1% /rpool
rpool/VARSHARE/zones 67G 31K 30G 1% /system/zones
rpool/VARSHARE/kvol 67G 31K 30G 1% /var/share/kvol
rpool/VARSHARE/sstore
67G 279M 30G 1% /var/share/sstore/repo
rpool/VARSHARE/tmp 67G 9.5M 30G 1% /var/share/tmp
rpool/VARSHARE/pkg 67G 32K 30G 1% /var/share/pkg
rpool/VARSHARE/pkg/repositories
67G 31K 30G 1% /var/share/pkg/repositories

For more information, see the df(8) man page.

Files Using the Most Space in a File System

Sorting files on a file system by their size and looking for the largest file might not be accurate.
Some file systems might contain files larger than the actual size occupied on the device. Such
files might seem very large though it occupies only a few blocks. To find the number of blocks
allocated for each file, use the disk usage (du) tool.

# du [-sh] [path ...]

The disk usage tool (du) checks all the files in a subtree, displays the number of blocks
allocated for each file, and ends the list with a total sum of blocks. If no path is specified, the du
tool performs its operation in the current working directory.

Use the -s option with the du command to display only the total sum for the subtree. The -h
option changes the sizes to human readable format.

Chapter 3 • Managing Data and Free Space 65

Quotas in File System

EXAMPLE 17 Displaying Directory Space Usage Summary

This example shows the output of the du command with total number of occupied blocks of all
the directories under the /usr subtree.
# du -sh /usr/*
86K /usr/X11
30M /usr/apache2
146M /usr/bin
0K /usr/dict
44M /usr/gnu
49M /usr/include
0K /usr/java
157M /usr/jdk
1.9M /usr/kernel
1.3G /usr/lib
9K /usr/proc
0K /usr/pub
938K /usr/sadm
56M /usr/sbin
63K /usr/sfw
726M /usr/share
3.8M /usr/xpg4
914K /usr/xpg6
452K /usr/xpg7

Quotas in File System

Oracle Solaris provides a mechanism to limit the amount of space a user or a group can occupy
in a file system. This mechanism is called quotas. Traditionally, quotas enable you to limit the
number of files or the number of blocks a user can allocate on a given file system.

Quotas are supported by UFS and ZFS file systems. NFS does not provide quotas by itself, but
can propagate the information from server local file system to client.

This section focuses on UFS quotas. For information about ZFS quotas, see “Setting ZFS
Quotas” in Managing ZFS File Systems in Oracle Solaris 11.4.

UFS quotas limit the number of files and number of allocated blocks per user. Quotas can either
be soft or hard. The soft and hard quota limits are applied to a user. When you exceed the soft
quota limit, you will receive a notification but you can add more files and allocate blocks for a
limited amount of time. When the time expires, you must remove the files and free the blocks
until the quota limit is met. When using hard quota, the user actions are limited immediately
when the limit is reached, and the response of the file system is identical to running out of
space.

66 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Quotas in File System

To activate quotas on a file system use the following commands.

# touch /tank/quotas
# quotaon /tank

Quotas are disabled automatically on mount or reboot. To disable quotas manually, use the
quotaoff command.

# quotaoff mountpoint

The quotaoff command disables quotas, and enables you to create files and allocate blocks
without limits.

You must enable quotas manually for each file system. By default, no limits are imposed on the
file system when you enable the quotas. Once the quotas are enabled, administrators can set
limits for selected users using the edquota command.

# edquota username

When you run the edquota command, it invokes an editor with line:

fs /tank blocks (soft = 0, hard = 0) inodes (soft = 0, hard = 0)

After the number is set for soft and hard quotas on the allocated blocks and files, the files
are edited and saved. Once saved, the quotas are automatically applied for the specified user
account.

The soft quota expiration timeouts can be set separately for each file and block using the
edquota -t command. For example:

# edquota -t

When you run the edquota command, it invokes an editor with line:

fs /tank blocks time limit = 1 week, files time limit = 7 days

The timeout can be specified in months, weeks, days, hours, minutes, or seconds. By default,
the timeout for hard and soft quotas are 7 days.

Quota checking can be disabled while users still have access to the file system. This might
cause inconsistency in quota counting. To synchronize the quota accounting with current state
of the file system, use the quotacheck command. For example:

# quotacheck mountpoint | block_device

As quotacheck reads the data directly from the device bypassing the kernel, any activity going
through the kernel is not accounted. If the file system cannot be unmounted for the duration

Chapter 3 • Managing Data and Free Space 67

Quotas in File System

of the command, it is highly suggested to turn off logging and minimize activities on the file
system.

You can also control access to the files by using file attributes and permissions. Setting
permissions and file attributes allow you to protect legitimate files. You can also use BART
to detect security breaches. For more information about data security and file integrity, see
Securing Files and Verifying File Integrity in Oracle Solaris 11.4.

For more information, see the edquota(8), quotacheck(8), and quotaoff(8) man pages.

68 Managing File Systems in Oracle Solaris 11.4 • November 2020

 ♦ ♦ ♦
4
C H A P T E R 4

File System Backup and Restore

This chapter describes procedures for generic backup and file system specific backup of data.

Backing Up and Restoring UFS File Systems

This section describes how to backup and restore a UFS file system using the ufsdump and
ufsrestore commands. This section also describes how to copy files and file systems using the
cpio, tar, and pax commands.

How the ufsdump Command Works

The ufsdump command performs two passes to back up a file system. During the first pass, this
command scans the raw device file for the file system, builds a table of directories and files
in the memory, and writes the table to a backup media. During the second pass, the ufsdump
command goes through the inodes in numerical order, reads the file contents and writes the data
to the backup media.

The ufsdump command writes a sequence of fixed-size records. When the ufsdump command
receives a notification that a record was partially written, the command assumes that it has
reached the physical end of a media. This method works for most devices. If a device is not able
to notify the ufsdump command that only a partial record has been written, a media error occurs
as the ufsdump command tries to write another record.

The ufsdump command automatically detects the end-of-media for most devices. Therefore, you
do not usually need to use the -c, -d, -s, and -t options to perform multi-volume backups.

You need to use the end-of-media options when the ufsdump command does not understand the
way the device detects the end-of-media.

Chapter 4 • File System Backup and Restore 69

Backing Up and Restoring UFS File Systems

To ensure compatibility with the ufsrestore command, the size option can still force the
ufsdump command to go to the next removable media before reaching the end of the current
media.

Copying Data Using the ufsdump Command

The ufsdump command copies data only from the raw disk slice. If the file system is still active,
any data in memory buffers might not be copied. The backup done by the ufsdump command
neither copies for free blocks, nor makes an image of the disk slice. If symbolic links point to
files on other slices, the link itself is copied. For more information, see the ufsdump(8) man
page.

Purpose of the /etc/dumpdates File

You can use the -u option with the ufsdump command to maintain and update the /etc/
dumpdates file. Each line in the /etc/dumpdates file shows the following information:
■ Backed up file system
■ Dump level of the last backup
■ Day, date, and time of the backup

A sample of the /etc/dumpdates file is as follows:

# cat /etc/dumpdates
/dev/rdsk/c0t0d0s0 0 Wed Jul 7 13:36:17 2016
/dev/rdsk/c0t0d0s7 0 Thu Jul 8 12:53:12 2016
/dev/rdsk/c0t0d0s7 9 Thu Jul 8 13:41:48 2016

When performing an incremental backup, the ufsdump command checks the /etc/dumpdates
file to find the date of the most recent backup of the next lower dump level. The command then
copies all the files that were modified since the date of that lower-level backup to the media.
After the backup is complete, a new information line which describes the backup you just
completed, replaces the information line for the previous backup at that level.

Use the /etc/dumpdates file to verify that backups are performed. This verification is
particularly important if you are having equipment problems. If a backup cannot be completed
because of an equipment failure, the backup is not recorded in the /etc/dumpdates file.

If you want to restore an entire disk, check the /etc/dumpdates file for a list of the most recent
dates and levels of backups so that you can determine in which tape you need to restore the
entire file system.

70 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Backing Up and Restoring UFS File Systems

Caution - The /etc/dumpdates file is a text file that can be edited. However, edit it only at your
own risk. If you make changes to the file that do not match your archive tapes, you might be
unable to find the tapes (or files) you need.

Backup Device Using the dump_file Argument

The dump_file argument (to the -f option) specifies the destination of the backup. The
destination can be one of the following devices:

■ Local tape drive

■ Remote tape drive
■ Standard output

Use this argument when the destination is not the default local tape drive, /dev/rmt/0. If you
use the -f option, then you must specify a value for the dump_file argument.

Note - The dump_file argument can also point to a file on a local disk or on a remote disk. If
done by mistake, this usage can fill up a file system.

Local Tape Drive Backup

Typically, the dump_file argument specifies a raw device file for a tape device. When the
ufsdump command writes to an output device, it creates a single backup file that might span
multiple tapes.

You can specify a tape device on your system by using a device abbreviation. The first device is
always 0. For example, if you have a LTO7 tape connected to a SAS HBA, use the device name,
/dev/rmt/0.

When you specify a tape device name, you can also type the letter n at the end of the name to
indicate that the tape drive should not rewind after the backup is complete. For example, /dev/
rmt/0n.

Use the -no-rewind option if you want to put more than one file onto the tape. If you run out of
space during a backup, the tape does not rewind before the ufsdump command asks for a new
tape.

Chapter 4 • File System Backup and Restore 71

Backing Up and Restoring UFS File Systems

Remote Tape Drive Backup

You can specify a remote tape device by using the syntax host:device. The ufsdump command
writes to the remote device when superuser on the local system has access to the remote
system. If you usually run the ufsdump command as a superuser, the name of the local system
must be included in the /.rhosts file on the remote system. If you specify the device as
user@host:device, the ufsdump command tries to access the device on the remote system as
the specified user. In this case, the specified user must be included in the /.rhosts file on the
remote system.

Using Standard Output With the ufsdump Command

When you specify a dash (-) as the dump_file argument, the ufsdump command writes to
standard output.

Note - The -v option (verify) cannot be used to verify a dump to standard output.

In a pipeline, you can use the ufsdump command to copy a file system by writing to a standard
output, and use the ufsrestore command to read from a standard input. For example:
# ufsdump 0f - /dev/rdsk/c0t0d0s7 | (cd /home; ufsrestore xf -)

Specifying Files to Back Up

You must always include file names as the last argument on the command line. This argument
specifies the source or content of the backup.

For a file system, specify the raw device file as /dev/rdsk/c0t0d0s7.

You can specify the file system by its mount point directory (for example, /export/home), as
long as an entry for it exists in the /etc/vfstab file.

Note - When you use the ufsdump command to back up one or more directories or files (rather
than a complete file system), a level 0 backup is done. Incremental backups do not apply.

Specifying Tape Characteristics

If you do not specify any tape characteristics, the ufsdump command uses a set of defaults. You
can specify the tape cartridge (c), density (d), size (s), and number of tracks (t). Note that you

72 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Specifying the ufsdump Command Options and Arguments

can specify the options in any order, as long as the arguments that follow match the order of the
options.

Limitations of the ufsdump Command

The ufsdump command has the following limitations:

■ Automatically calculates the number of tapes that are needed for backing up file systems.
You can use the dry run mode (-S option) to determine how much space is needed before
actually backing up file systems.
■ Provides built-in error checking to minimize problems when it backs up an active file
system.
■ Backs up files that are remotely mounted from a server. Files on the server must be backed
up on the server itself. Users are denied permission to run the ufsdump command on files
that they own, which are located on a server.

Specifying the ufsdump Command Options and Arguments

This section describes the options and arguments for the ufsdump command. The syntax for the
ufsdump command is as follows:

/usr/sbin/ufsdump options arguments filenames

options Single string of one-letter option names.

arguments Identifies option arguments and might consist of multiple strings. The
option letters and their associated arguments must be in the same order.

filenames Identifies the files to back up. These arguments must always come last,
each separated by a space.

Default ufsdump Options

You can run the ufsdump without any options by using the following syntax:

# ufsdump filenames

Chapter 4 • File System Backup and Restore 73

Specifying ufsrestore Options and Arguments

The ufsdump command uses the following options and arguments, by default:

# ufsdump 9uf /dev/rmt/0 filenames

The option, 9in the ufsdump command will do a level 9 incremental backup of all the modified
files to the default tape drive since the previous backup.

The -u option is used to update the /etc/dumpdates file with information about the current
backup. The /etc/dumpdates contains information about the previous backup and can decide
which are the files to be backed-up.

The f /dev/rmt/0 option sends backup to the first tape device directly connected to an Oracle
Solaris system.

For more information, see the ufsdump(8) man page.

The ufsdump Command and Security Issues

Consider the following aspects to ensure that there are no security issues with the file system:

■ Require superuser access to use the ufsdump command.

■ If you are performing centralized backups, ensure that superuser access entries are removed
from the /.rhosts files on the clients and servers.

For more information, see Oracle Solaris 11.4 Security and Hardening Guidelines.

Specifying ufsrestore Options and Arguments

The syntax of the ufsrestore command is as follows:

# /usr/sbin/ufsrestore options arguments filenames

options Single string of one-letter option names. You must include only one
option.

arguments Follows the option string with the arguments that match the options. The
option letters and their associated arguments must be in the same order.

filenames Specifies the file or files to be restored as arguments to the -x or -t

options. These arguments must always come last, separated by spaces.

74 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Copying Files and File System

For more information, see the ufsrestore(8) man page.

Copying Files and File System

This section includes information about the commands that you can use to copy files and file
systems. It also contains information about how to copy files using different commands.

Commands for Copying File Systems

If you want to copy or move individual files, portions of file systems, or complete file systems,
follow the procedures described in this section.

The following table describes the various backup and restore commands that are available in
Oracle Solaris. For enterprise environments, consider using an enterprise-level backup product.

TABLE 1 Advantages and Disadvantages of tar, pax, and cpio Commands

Command Function Advantages Disadvantages

tar Copies files and directory ■ Available on most UNIX ■ Is not aware of file system
subtrees to a single tape. operating systems boundaries
■ Public domain versions are ■ Length of full path
readily available name cannot exceed 255
characters
■ Cannot be used to create
multiple tape volumes
pax Copies files, special files, ■ Better portability than the Same disadvantages as the tar
or file systems that require tar or cpio commands for command, except that the pax
multiple tape volumes. Or, POSIX-compliant systems command can create multiple
copies files to and from ■ Multiple vendor support tape volumes.
POSIX-compliant systems.
cpio Copies files, special files, ■ Packs data onto tape more The command syntax is more
or file systems that require efficiently than the tar difficult than the tar or pax
multiple tape volumes. Or, command commands.
copies files from systems ■ Skips over any bad spots in a
running latest Oracle Solaris tape when restoring
version to systems running
■ Provides options for writing
older Oracle Solaris versions.
files with different header
formats, such as tar, ustar,
crc, odc, bar, for portability
between different system
types

Chapter 4 • File System Backup and Restore 75

How to Copy Directories Between File Systems Using the cpio Command

Command Function Advantages Disadvantages

■ Creates multiple tape
volumes

The following sections describes step-by-step instructions and examples of how to use these
commands.

Copying Directories Between File Systems Using

cpio Command

You can use the cpio (copy in and out) command to copy individual files, groups of files, or
complete file systems. This section describes how to use the cpio command to copy complete
file systems.

The cpio command is an archiving program that copies a list of files into a single, large output
file. This command inserts headers between the individual files to facilitate recovery. You can
use the cpio command to copy complete file systems to another slice, another system, or to a
media device, such as tape.

Because the cpio command recognizes end-of-media and prompts you to insert another
volume, it is the most effective command, other than ufsdump, to create archives that require
multiple media.

With the cpio command, you can use the ls and find commands to list and select the files you
want to copy, and then to pipe the output to the cpio command.

How to Copy Directories Between File Systems Using the cpio

Command
1. Change to the appropriate directory.

# cd filesystem1

2. Copy the directory tree from filesystem1 to filesystem2 by using a combination of the
find and cpio commands.

# find . -print -depth | cpio -pdm filesystem2

-print Prints the file names

76 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Copy Directories Between File Systems Using the cpio Command

-depth Descends the directory hierarchy and prints file names from the bottom
up

-p Creates a list of files

-d Creates directories as needed

-m Sets the correct modification times on directories

The files from the directory name that is specified are copied. The symbolic links are preserved.
You can also specify the -u option to perform unconditional copy. Otherwise, older files do not
replace newer files. This option might be useful if you want an exact copy of a directory, and
some of the files being copied might already exist in the target directory.
For more information, see the cpio(1) man page.

3. Verify that the copy was successful by displaying the contents of the destination
directory.
# cd filesystem2
# ls

4. If required, remove the source directory.

# rm -rf filesystem1

Example 18 Copying Directories Between File Systems Using the cpio Command

This example shows how to copy directories between file systems using the cpio command.
# cd /data1
# find . -print -depth | cpio -pdm /data2
19013 blocks
# cd /data2
# ls
# rm -rf /data1

Copying Files and File Systems to a Tape Device

You can use the tar, pax, and cpio commands to copy files and file systems to a tape device.
You can choose the command based on the amount of flexibility and precision you require for
the copy. Because all three commands use the raw device, you do not need to format or create a
file system on tapes before you use them. The tape drive and device name that you use depend

Chapter 4 • File System Backup and Restore 77

How to Copy Files to a Tape Device Using the tar Command

on the hardware configuration for each system. For more information about tape devices, see
“Choosing Which Media to Use” in Managing Devices in Oracle Solaris 11.4.

Copying Files to a Tape Device Using the tar Command

Review the following behaviors before you copy files to tape using the tar command:
■ Copying files to a tape with the -c option to the tar command destroys any files already on
the tape, at or beyond the current tape position.
■ You can use file name substitution wildcards (? and *) as a part of the file name that you
specify when copying files. For example, to copy all documents with .doc extension, type
*.doc as the file name argument.
■ You cannot use file name substitution wildcards when you extract files from a tar archive.

For more information, see the tar(1) man page.

How to Copy Files to a Tape Device Using the tar Command

1. Change to the directory that contains the files you want to copy.

2. Insert a write-enabled tape into the tape drive.

3. Copy the files to the tape device.

$ tar cvf /dev/rmt/n filenames

c Creates an archive.

v Displays the name of each file as it is archived.

f /dev/rmt/n Indicates that the archive should be written to the specified device or file.

filenames Indicates the files and directories that you want to copy. You can copy
multiple files by specifying the file names with spaces.
The file names that you specify are copied to the tape, overwriting any
existing files on the tape.

4. Remove the tape from the drive. Write the names of the files on the tape label.

5. Verify that the files are copied to the tape.

$ tar tvf /dev/rmt/n

78 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to List All the Files on a Tape Device Using the tar Command

Example 19 Copying Files to a Tape Device Using the tar Command

The following example shows how to copy three files to the tape in tape drive 0 using the tar
command.

$ cd /export/home/kryten
$ ls reports
reportA reportB reportC
$ tar cvf /dev/rmt/0 reports
a reports/ 0 tape blocks
a reports/reportA 59 tape blocks
a reports/reportB 61 tape blocks
a reports/reportC 63 tape blocks
$ tar tvf /dev/rmt/0

How to List All the Files on a Tape Device Using the tar Command

1. Insert a tape device into the tape drive.

2. Display the tape contents.

$ tar tvf /dev/rmt/n

-t Lists the table of contents for the files on the tape.

-v Provides detailed information about the files on the tape, used with the -t
option.

-f /dev/rmt/n Indicates the tape device.

Example 20 Listing the Files on a Tape Device Using the tar Command

The following example shows a listing of files on the tape in drive 0.

$ tar tvf /dev/rmt/0

drwxr-xr-x 0/0 0 Jul 14 13:50 2010 reports/
-r--r--r-- 0/0 206663 Jul 14 13:50 2010 reports/reportC
-r--r--r-- 0/0 206663 Jul 14 13:50 2010 reports/reportB
-r--r--r-- 0/0 206663 Jul 14 13:50 2010 reports/reportA

How to Retrieve Files From a Tape Device Using the tar Command

1. Change to the directory where you want to save the retrieved files.

Chapter 4 • File System Backup and Restore 79

How to Retrieve Files From a Tape Device Using the tar Command

2. Insert the tape into the tape drive.

$ tar xvf /dev/rmt/n [filenames]

-x Indicates that the files must be extracted from the specified archive
file. All files on the tape in the specified drive are copied to the current
directory.

-v Displays the name of each file as it is retrieved.

-f /dev/rmt/n Indicates the tape device that contains the archive.

filenames Specifies a file to retrieve. You can retrieve multiple files by specifying
the file names with spaces.

3. Verify that the files are copied.

$ ls -l

Example 21 Retrieving Files From a Tape Device Using the tar Command

The following example shows how to retrieve all the files from the tape in drive 0 using the tar
command.

$ cd /var/tmp
$ tar xvf /dev/rmt/0
x reports/, 0 bytes, 0 tape blocks
x reports/reportA, 0 bytes, 0 tape blocks
x reports/reportB, 0 bytes, 0 tape blocks
x reports/reportC, 0 bytes, 0 tape blocks
x reports/reportD, 0 bytes, 0 tape blocks
$ ls -l

The names of the files extracted from the tape must exactly match the names of the files that are
stored on the archive. If you have any doubts about the names or paths of the files, first list the
files on the tape.

For more information, see the tar(1) man page.

Copying Files to a Tape Device Using the pax Command

This section describes how to copy files to a tape device using the pax command.

80 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Copy Files to a Tape Device Using the pax Command

How to Copy Files to a Tape Device Using the pax Command

1. Change to the directory that contains the files you want to copy.

2. Insert a write-enabled tape into the tape drive.

3. Copy the files to the tape.

$ pax -w -f /dev/rmt/n filenames

-w Enables the write mode.

-f /dev/rmt/n Identifies the tape drive.

filenames Indicates the files and directories that you want to copy. You can copy
multiple files by specifying the file names with spaces.

4. Verify that the files are copied to the tape device.

$ pax -f /dev/rmt/n

5. Remove the tape device from the drive. Write the names of the files on the tape
label.

Example 22 Copying Files to a Tape Device Using the pax Command

The following example shows how to use the pax command to copy all the files in the current
directory.
$ pax -w -f /dev/rmt/0 .
$ pax -f /dev/rmt/0
filea fileb filec

Copying Files to a Tape Device Using the cpio Command

This section describes how to copy files to a tape device using the cpio command.

How to Copy All the Files in a Directory to a Tape Device Using the cpio
Command

1. Change to the directory that contains the files you want to copy.

2. Insert a write-enabled tape into the tape drive.

Chapter 4 • File System Backup and Restore 81

How to Copy All the Files in a Directory to a Tape Device Using the cpio Command

3. Copy the files to the tape device.

$ ls | cpio -oc > /dev/rmt/n

ls Provides the cpio command with a list of file names.

-o Specifies that the cpio command should operate in copy-out mode. This
option ensures portability to other systems of the vendors.

-c Specifies that the cpio command must write header information in ASCII
character format. This option ensures portability to other systems of the
vendors.

> /dev/rmt/n Specifies the output file.

All the files in the directory are copied to the tape in the drive you specify, overwriting any
existing files on the tape. The total number of blocks that are copied is displayed in the output.

4. Verify that the files are copied to the tape device.

$ cpio -civt < /dev/rmt/n

-c Specifies that the cpio command must read files in ASCII character
format.

-i Specifies that the cpio command must operate in copy-in mode, even
though the command is only listing files at this point.

-v Displays the output in a format that is similar to the output from the ls -
l command.

-t Lists the table of contents for the files on the tape in the tape drive that
you specify.

< /dev/rmt/n Specifies the input file of an existing cpio archive.

5. Remove the tape from the drive. Write the names of the files on the tape label.

Example 23 Copying All the Files in a Directory to a Tape a Device Using the cpio Command

The following example shows how to copy all the files in the /export/home/kryten directory
to a tape device in the tape drive 0.

$ cd /export/home/kryten

82 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to List the Files on a Tape Device Using the cpio Command

$ ls | cpio -oc > /dev/rmt/0

1280 blocks
$ cpio -civt < /dev/rmt/0
-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, filea
-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, fileb
-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, filec
drwxr-xr-x 2 kryten staff 0 Jul 14 13:52 2010, letters
drwxr-xr-x 2 kryten staff 0 Jul 14 13:52 2010, reports
1280 blocks

For more information, see the cpio(1) man page.

How to List the Files on a Tape Device Using the cpio

Command

This procedure shows how to list the files on a tape device using the cpio command.

Note - Listing the table of contents on a tape takes a long time because the cpio command must
process the entire archive.

1. Insert an archive tape into the tape drive.

2. List the files on the tape.

$ cpio -civt < /dev/rmt/n

Example 24 Listing the Files on a Tape Device Using the cpio Command

The following example shows how to list the files on the tape in drive 0.

$ cpio -civt < /dev/rmt/0

-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, filea
-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, fileb
-r--r--r-- 1 kryten staff 206663 Jul 14 13:52 2010, filec
drwxr-xr-x 2 kryten staff 0 Jul 14 13:52 2010, letters
drwxr-xr-x 2 kryten staff 0 Jul 14 13:52 2010, reports
1280 blocks

Chapter 4 • File System Backup and Restore 83

How to Retrieve All the Files From a Tape Device Using the cpio Command

How to Retrieve All the Files From a Tape Device Using the
cpio Command

If the archive is created using relative path names, the input files are built as a directory within
the current directory when you retrieve the files. If the archive is created with absolute path
names, the same absolute paths are used to recreate the file on your system.

Caution - The use of absolute path names can be dangerous because you might overwrite
existing files on your system.

1. Change to the directory where you want to save the retrieved files.

2. Insert the tape into the tape drive.

3. Extract all the files from the tape device.

$ cpio -icvd < /dev/rmt/n

-i Extracts files from standard input.

-c Specifies that the cpio command must read files in ASCII character
format.

-v Displays the files as they are retrieved in a format that is similar to the
output from the ls command.

-d Creates directories as needed.

< /dev/rmt/n Specifies the output file.

4. Verify that the copy was successful by displaying the contents of the destination
directory.

$ ls -l

Example 25 Retrieving All the Files From a Tape Device Using the cpio Command

The following example shows how to retrieve all the files from the tape in drive 0.

$ cd /var/tmp
$ cpio -icvd < /dev/rmt/0
answers
sc.directives

84 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Retrieve a Specific File From a Tape Device Using the cpio Command

tests
8 blocks
$ ls -l

How to Retrieve a Specific File From a Tape Device Using the

cpio Command

This procedure shows how to retrieve a specific file from a tape device using the cpio
command.

1. Change to the directory where you want to save the retrieved files.

2. Insert the tape into the tape drive.

3. Retrieve a subset of files from the tape.

$ cpio -icv "*file" < /dev/rmt/n

-i Extracts files from standard input.

-c Specifies that the cpio command should read headers in ASCII character
format.

-v Displays the files as they are retrieved in a format that is similar to the
output from the ls command.

*file Specifies that all files that match the pattern are copied to the current
directory. You can specify multiple patterns, but each pattern must be
enclosed in double quotation marks.

< /dev/rmt/n Specifies the input file.

4. Verify that the files are copied.

$ ls -l

Example 26 Retrieving a Specific File From a Tape Device Using the cpio Command

The following example shows how to retrieve all files with the chapter suffix from the tape in
drive 0.

$ cd /home/smith/Book
$ cpio -icv "*chapter" < /dev/rmt/0

Chapter 4 • File System Backup and Restore 85

How to Copy Files to a Remote Tape Device Using tar and dd Command

Boot.chapter
Directory.chapter
Install.chapter
Intro.chapter
31 blocks
$ ls -l

For more information, see the cpio(1) man page.

Copying Files and File Systems to a Remote Tape

Device
This section shows how to copy files and file systems to a remote tape device.

How to Copy Files to a Remote Tape Device Using tar and dd

Command
1. Configure ssh on the remote system so that you can access the tape drive.
For more information about configuring ssh, see Managing Secure Shell Access in Oracle
Solaris 11.4.

2. Change to the directory where you want to put the files.

3. Insert the tape into the tape drive.

4. Copy the files to a remote tape drive.

$ tar cvf - filenames | ssh remote-host dd of=/dev/rmt/n obs=block-size

-cf Creates a tape archive, lists the files as they are archived, and specifies
the tape device.

-v Provides additional information about the tar file entries.

- (Hyphen) Represents a placeholder for the tape device.

filenames Identifies the files to be copied. You can copy multiple files by specifying
the file names with spaces.

ssh | remote- Pipes the output of the tar command to a remote system.
host

86 Managing File Systems in Oracle Solaris 11.4 • November 2020

 How to Extract Files From a Remote Tape Device

dd of= /dev/rmt/ Represents the output device.

n

obs=block-size Represents the blocking factor.

5. Remove the tape from the drive. Write the names of the files on the tape label.

Example 27 Copying Files to a Remote Tape Drive Using the tar and dd Commands

# tar cvf - * | ssh mercury dd of=/dev/rmt/0 obs=126b

password:
a answers/ 0 tape blocks
a answers/test129 1 tape blocks
a sc.directives/ 0 tape blocks
a sc.directives/sc.190089 1 tape blocks
a tests/ 0 tape blocks
a tests/test131 1 tape blocks
6+9 records in
0+1 records out

How to Extract Files From a Remote Tape Device

1. Insert the tape into the tape drive.

2. Change to a temporary directory.

$ cd /var/tmp

3. Extract the files from a remote tape device.

$ ssh remote-host dd if=/dev/rmt/n | tar xvBpf -

ssh remote-host Indicates a secure shell that is started to extract the files from the tape
device by using the dd command.

dd if=/dev/rmt/n Indicates the input device.

| tar xvBpf - Pipes the output of the dd command to the tar command, which is used
to restore the files.

4. Verify that the files have been extracted.

$ ls -l

Chapter 4 • File System Backup and Restore 87

How to Extract Files From a Remote Tape Device

Example 28 Extracting Files From a Remote Tape Device

$ cd /var/tmp
$ ssh mercury dd if=/dev/rmt/0 | tar xvBpf -
password:
x answers/, 0 bytes, 0 tape blocks
x answers/test129, 48 bytes, 1 tape blocks
20+0 records in
20+0 records out
x sc.directives/, 0 bytes, 0 tape blocks
x sc.directives/sc.190089, 77 bytes, 1 tape blocks
x tests/, 0 bytes, 0 tape blocks
x tests/test131, 84 bytes, 1 tape blocks
$ ls -l

For more information, see the tar(1) and dd(8) man pages.

88 Managing File Systems in Oracle Solaris 11.4 • November 2020

 ♦ ♦ ♦
5
C H A P T E R 5

Managing File System Performance

This chapter describes how to manage the file system performance using commands such as
fsstat and tunefs.

Monitoring File System Performance Using fsstat

Starting with Oracle Solaris 11, you can use the fsstat command to report file system
operations. You can report file system operations in multiple ways. For example, reports based
on mount point or file system type.

For example, use the fsstat command to display all ZFS file system operations since the ZFS
module is loaded:

$ fsstat zfs
new name name attr attr lookup rddir read read write write
file remov chng get set ops ops ops bytes ops bytes
268K 145K 93.6K 28.0M 71.1K 186M 2.74M 12.9M 56.2G 1.61M 9.46G zfs

For example, use the fsstat command to display all file system operations since the /export/
ws file system is mounted:

$ fsstat /export/ws
new name name attr attr lookup rddir read read write write
file remov chng get set ops ops ops bytes ops bytes
0 0 0 18.1K 0 12.6M 52 0 0 0 0 /export/ws

The default form is to report statistical information in easy to understand values, such as GB,
KB, and MB. For more information, see the fsstat(8) man page.

Chapter 5 • Managing File System Performance 89

Tuning UFS File System Performance Using tunefs

Tuning UFS File System Performance Using tunefs

UFS provides the tunefs tool which enables you to modify some pf the performance related
parameters of a file system. The parameters that can be modified are:
■ The maximum number of logical blocks, belonging to one file, that is allocated
contiguously. Contiguously allocated regions enable more data to be transferred in one I/O
request. The value can be set to any positive integer number. The actual value will be the
lesser of what has been specified and what the hardware supports.
■ The maximum number of contiguous logical blocks any single file can allocate from a
cylinder group.
■ The minimum free space threshold, or the percentage of space held back from normal users.

For more information, see “Customizing UFS File System Parameters” on page 19 and the
tunefs(8) man page.

UFS Direct I/O Mode

File systems usually perform a lot of caching to avoid redundant I/O access to a device. This
boosts performance when a block of data is needed multiple times over a short period of
time. However, for some loads, the file system caching might not be effective or can even be
disruptive. This usually involves applications (mostly databases) which perform caching on
their own, and decide what needs to be cached in the memory for faster access. For such loads,
UFS provides direct I/O mode, which disables caching on the file system level, enabling all
read and write operations to be performed directly on the device. UFS then remains in control
of the file system organization but data flow is managed by the application.

Direct I/O mode can be enabled for a whole file system or for each file. To enable direct I/O
mode for a file system, you can use the directio option while mounting the file system. Direct
I/O access is then used to access all the files on the mounted file system.

# mount -F UFS -o directio /dev/dsk/device-name/directory-name

Enabling direct I/O mode on separate files is useful on file systems with mixed load. To enable
direct I/O access for a specific file, you must update the source code of the application when the
file is opened. For more information, see the directio(3C) and mount_ufs(8) man pages.

90 Managing File Systems in Oracle Solaris 11.4 • November 2020

Index

Numbers and Symbols /proc, 14

4.3 Tahoe file system, 12 /tmp, 14
Disk based file system, 25
disk-based file systems, 12
DOS
B file system, 12
BSD Fat Fast File system, 12

E
C /etc/vfstab file, 28
CD or DVD /etc/vfstab, 26
creating file system, 23 /etc/vfstab file
copy directories between file systems field description, 29
cpio command, 76 extended file attributes
copy files to a tape runat command, 15
tar command, 78
copying file systems, 75
copying files to tape
pax command, 81 F
cpio command file systems
copying directories, 76 4.3 Tahoe, 12
creating file system, 17 BSD Fat Fast, 12
CTFS file system, 13 CTFS, 13
disk-based, 12
DOS, 12
finding types, 24
D High Sierra, 12
default ISO 9660, 12
file system for /tmp (TMPFS), 14 network-based, 13
determining OBJFS, 14
file system types, 24 PCFS, 12
direct I/O mode, 90 process, (overview), 14
directio option, 90 PROCFS, (overview), 14
directories pseudo, (overview), 13

91
Index

SHAREFS, 14 automount command, 27

stopping all processes accessing (how to), 34 network-based file systems, 13
TMPFS, 14 newfs command, 18, 19
UFS, 12
UNIX, 12
ZFS, 12, 12
finding O
type of file system, 24 OBJFS file system, 14
fsstat command, 89

P
H PCFS
High Sierra file system, 12 creating file system, 22
HSFS See High Sierra file system PCFS file system, 12, 26
/proc directory, 14
process file system (PROCFS), 14
I PROCFS file system
ISO 9660 file system, 12 (overview), 14
pseudo file systems
(overview), 13

K
killing
all processes accessing a file system (how to), 34 R
retrieving files on a tape
tar command, 79
Rock Ridge extension (HSFS file system), 12
L
listing files on a tape
tar command, 79
S
SHAREFS file system, 14
M stopping
mkfs command, 17 all processes for a file system (how to), 34
mkfs_ufs command, 19 system performance
mkisofs command, 23 fsstat command, 89
MNTFS file system, 14 tunefs command, 90
mnttab file system, 14
mount command, 25
T
tar command
N copy files to a tape device, 78
Network based file system, 27 /tmp directory, 14

92 Managing File Systems in Oracle Solaris 11.4 • November 2020

 Index

TMPFS file system

overview, 14
tunefs command, 90

U
UFS
creating file system, 18
creating multi-terabyte file system, 19
UFS file system, 12
UFS file system parameter
fragment size, 20
logical block size, 19
minimum free space, 21
number of inodes, 21
optimization type, 21
ufsdump command, 69
UNIX file system, 12

V
Virtual File System Table, 28

Z
ZFS file system, 12, 12

93
94 Managing File Systems in Oracle Solaris 11.4 • November 2020