NBMR AIX QuickStartGuide
NBMR AIX QuickStartGuide
May 2017
Version 8.1.1
The software contains proprietary information of Cristie Software Ltd.; it is provided under a license agreement
containing restrictions on use and disclosure and is also protected by copyright law. Reverse engineering of the
software is prohibited.
Due to continued product development this information may change without notice. The information and intellectual
property contained herein is confidential between Cristie Software Ltd. and the client and remains the exclusive
property of Cristie Software Ltd. If you find any problems in the documentation, please report them to us in writing.
Cristie Software Ltd. does not warrant that this document is error-free.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise without the prior written permission of Cristie
Software Ltd.
IBM Tivoli Storage Manager (TSM), AIX and TIVOLI are trademarks of the IBM Corporation.
Amazon Web Services (AWS) and Amazon Elastic Compute Cloud (EC2) are trademarks of Amazon.com, Inc.
Cristie Management Console (CMC), PC-BaX, UBax, Cristie Connect, Cristie Storage Manager (CSM), SDB, ABMR
(Bare Machine Recovery for EMC Avamar), NBMR (Bare Machine Recovery for EMC NetWorker), TBMR (Bare Machine
Recovery for Spectrum Protect/TSM), CBMR (Cristie Bare Machine Recovery), Recovery Simulator (RS) and CRISP
(Cristie Recovery ISO Producer) are all trademarks of Cristie Software Ltd.
Contents
2 Document Conventions 5
3 System Requirements 6
5 Installation 8
5.1 Install
...................................................................................................................................
via Smit/Smitty 8
5.2 Install
...................................................................................................................................
via RPM 8
5.3 License
................................................................................................................................... 8
5.4 Uninstall
................................................................................................................................... 9
6 Product Licensing 10
6.1 Trial
...................................................................................................................................
License 10
6.2 Full
...................................................................................................................................
License 11
6.2.1 Setting up.........................................................................................................................................................
a Cristie Licensing Portal account 11
6.2.2 Online Activation
......................................................................................................................................................... 13
6.2.3 Manual Activation
......................................................................................................................................................... 15
8 Performing a DR backup 19
8.1 Recording
...................................................................................................................................
System Information 19
8.2 Creating
...................................................................................................................................
Networker Backup 21
The process of backing up and recovering an AIX machine comprises three phases:
Stages 1 and 3 may be performed using the Graphical User Interface run from the command nbmr.
The disaster recovery backup must be performed using the EMC NetWorker backup tools - please
refer to your EMC NetWorker documentation for instructions on how to do this.
Note: NBMR must be installed and run by a user that has root access
2 Document Conventions
The following typographical conventions are used throughout this guide:
3 System Requirements
NBMR for AIX requires the following minimum hardware requirements:
Memory 1GB
NBMR for AIX is suitable for all versions of AIX after 6.1. However, ACLs (Access Control Lists) are
only supported in version 6.1 and above.
For the new Licensing Manager to access the machine to be licensed, Cristie Deployment Service
must be started from the command line by entering #cds
Prerequisites:
ofpthread64libs-1.0.0
Screen >=4.0.3
libxslt >=1.1.5
These are included with the installation and may also be downloaded from the IBM AIX Toolbox for
Linux website:
http://www-03.ibm.com/systems/power/software/aix/linux/toolbox/download.html
5 Installation
NBMR can be installed via the AIX System Management Interface - smit or smitty - directly via
RPM or from the archive. It is recommended that all installation files are installed using the same
method.
The installation of NBMR requires the RPM package management tool: fileset rpm.rte
3.0.5.20. This tool is installed by default on AIX ML6, 6.1 ML1 and AIX 7.1 and 7.2 MLI.
The version of this tool can be checked using the command 'lslpp -1 rpm.rte'. If the fileset is
at an earlier maintenance level, then the rpm.rte file can be downloaded individually from: http://
www-933.ibm.com/support/fixcentral/
The installation directory should be the '/bffs' directory on the CD or in the archive.
The prerequisites listed on the preceeding page are contained in the directory alongside NBMR and
are installed automatically.
Note: the smitty installation delegates to RPM. Therefore, if some packages have
already been installed via RPM then the latest version available is selected.
5.3 License
Following the instructions in this section will result in a standard 30-day trial license being installed.
Cristie provide a 30 day trial license so that the product can be fully evaluated before purchase.
If you have purchased a full license, you will have been sent a contract identifier and activation code,
these can be used to activate the product with the licmgr tool as follows:
Note the above codes are examples only - please use the activation codes sent to you. More
information about the licmgr tool can be found by typing 'man licmgr'.
5.4 Uninstall
To uninstall if installed via smit or smitty, run 'smit remove' or 'smitty remove', then select
the relevant packages for removal.
6 Product Licensing
When first installed, NBMR may be used for a trial period of 30 days. During that period NBMR is
fully functional. If the software is subsequently un-installed and later re-installed on the same
system, the 30 day period continues from the date of the first installation.
If you wish to use the software beyond the trial period, you must register and purchase a license
from Cristie Software Ltd.. Alternatively, and in special circumstances, Cristie Software Ltd. may
extend the license period if you wish to trial the software beyond that period.
If you purchase the product, then contract and license activation codes will be available on the
Cristie Licensing Portal. Together these codes will enable you to fully activate the product.
# licmgr -p nbmr
Entering this command, will display the Cristie License Manager, that shows Machine attributes,
Contract ID, the installed host System signature, the current Cristie product (NBMR in this case),
the product version, the trial end date and the licence Status.
The NBMR configuration file generator will become active again as soon as a full license has been
purchased from Cristie Software Ltd. and the new contract and activation codes entered via the
Cristie License Manager.
To upgrade from the trial license to a full license, you need to apply for a full license activation code
either via the Cristie Licensing Portal website or via the product Cristie License Manager. In either
case you will need to first register an account on the Cristie Licensing Portal (located at https://
portal.cristie.com/login). A Contract ID will be created and provided to you when you purchase a
license.
These are the various codes used in the Cristie licensing process:
Contract ID: A 4-digit number supplied by Cristie Software Ltd. Sales during the license purchase
process.
Agreement Number: Same as Contract ID at the moment.
Contract Code: 35-character contract code obtained from the Cristie Licensing Portal
Activation Code: 35-character support activation code obtained from the Cristie Licensing Portal
In special circumstances a 'bulk license' may be issued by Cristie Software Ltd. for customers that
order a significant number of product licenses. Please contact your Cristie sales representative if
you wish to discuss this service.
Note this discussion assumes that NBMR is already installed on a Customer production
machine.
To setup a new account on the Cristie Licensing Portal follow the following steps. To do this you will
need your 4-digit Contract ID and contract setup password. These will be provided by email from
Cristie Software Ltd. when you purchase a product license.
Note: Your Contract ID may have been supplied to you as your contract Agreement
Number. In that case please use your Agreement number in place of the Contract ID
throughout.
1. On a machine that has Internet access run a suitable browser such as Microsoft Internet Explorer
and navigate to the Cristie Licensing Portal web page at https://portal.cristie.com/login.
Select Register to create a new account. Enter your new account details (note this is an example):
At this point you may now log in to the Cristie Licensing Portal using the E-mail ID and password
setup in the previous steps.
1. Assign your Contract code on the NBMR host machine, by opening up a terminal and entering:
(The required email and password are the ones you use to sign into the Cristie Licensing Portal)
The Cristie License Manager will be refreshed showing your Contract ID, the new activation code and
your contract support end date.
Note: Internet access on the host machine is required to run the online activation
process from the Cristie License Manager directly.
The Cristie License Manager will be refreshed showing your Contract ID, the new activation code and
your contract support end date.
The first step is to create a recovery CD or PXE/NIM bootable image. This is an iso image that can
either be burned to CD to boot the recovery environment or be extracted to create a network boot
environment.
The terminal type can be specified here if a different terminal is desired on boot up. xterm is
recommended as it is compatible with most terminal emulation programs on Windows and UNIX.
The temporary directory is used to create the structure of the CD, which consequently is converted
to a file. The GUI creates an iso file which may be burned to a CD using an iso burning tool such as
burn_cd:
> burn_cd -d /dev/cd0 recovery_cd.iso
Create a DHCP/BOOTP entry for the machine with option 151 specifying the NFS server IP
address and 152 specifying the NFS server path
On an AIX NIM Master, the DHCP configuration itself ('/etc/dhcpsd.cnf') contains detailed
instructions to set up a host in the manner detailed above.
This is an example section of /etc/dhcpsd.cnf used to enable NIM booting of the recovery
environment for the machine with MAC address '01:02:03:04:05:06:07:08':
The attributes for the NFS server address and directory are set in a similar manner.
Alternatively, the NFS server and directory may be set statically for the image by entering the full
address (ie. 10.10.14.90:/mnt/SPOT) or just a path (ie. /mnt/SPOT) into the NFS Server of the
dialogue:
Note: if this option is used, then the recovery environment attempts to boot from the NFS
server and directory given. If only a path is supplied then the recovery environment
attempts to use the boot server (either BOOTP or DHCP) and the path supplied. See the
auxilliary document "AIXBootingProcedures.pdf" for a more complete explanation"
8 Performing a DR backup
Performing a DR backup is split into four stages:
The system information is recorded to allow the recovery environment to recreate the original system
environment. This will include drive and file-system information, as well as information about
essential packages for rebuilding the system (for example, to provide file encryption at recovery
time).
Selecting System Configuration from the main menu opens a sub-menu containing options for
creating the configuration:
The Options menu item includes Include SAN Devices: Yes/No. If this option is set to Yes then all
disks (including SAN attached disks) are parsed for inclusion in the configuration files. Only set this
options to No if you are certain that no volumes that you wish to recover are on SAN.
The default location of the configuration information is /NBMRCFG/disrec.xml.
When running the configuration tool information, the current operations are displayed:
Once this operation is complete, the log fine can be found in /var/log/cristie/nbmrcfg.log.
This may also be viewed using the Log Files submenu.
3. FileSystems - create file-systems on the logical volumes created in the previous step
Additional steps are required when Veritas Volume Manager is installed and Veritas Volume Groups
must be recovered, these are:
All stages are run though in order - consequently this can take a long time dependent upon the
speed of disks and network interfaces. Once the recovery is complete, the system can be rebooted
into its original state.
Before re-boot, however, it is very useful to make a copy of the log files generated during the
recovery as shown in Copying Log Files.
Once startup is complete, you must enter the username and password 'root' and 'root' to enter the
recovery environment. This is required to enable AIX multi-user support and allow the use of ctrl-c to
halt operations.
Note: if the 'screen' utility is installed, then this step is bypassed as screen provided its
own multi-user environment
The network setup must be configured with the details of the machine backed up to the Networker
server.
IP Address: 10.10.22.3
Hostname aix61xxxxxx
Subnet 255.0.0.0
Default Gateway 10.0.1.100
DNS 10.0.1.110
The above source machine values have been configured in the Networker Server hosts file and client
properties.
The setup network dialogue based on the above example is populated as below.
Note: if no DNS entry is given, then all subsequent addresses MUST be given in dotted
decimal form
This presents two recovery styles - automatic and manual - as well as tools for managing the
recovery environment and log files.
The automatic recovery runs through all stages of the recovery and only provides options to
The manual recovery allows the option of recovering only selected volume groups and
running selected phases of the recovery individually
Note: if the graphical environment is unusable at this stage, for example if the currently
selected item appears to change unexpectedly, then the terminal type should be
changed. See the Trouble-shooting section for further details
2. Backup Location - specify the attributes of the location containing the backup
5. Copy Log Files - copy the log files generated by the recovery
Setup Network
The network can be setup for any interfaces found using manual configuration. This is a requirement
for the NBMR recovery environment.
The manual configuration step is exactly the same as the initial network setup in Starting the
Recovery Environment.
Backup Location
The Backup Location menu allows the user to setup the NetWorker client for access to the server.
It is vitally important that the hostname and IP address of the client match that machine that was
backed up.
Note: if the server is specified by name, then a DNS server must have been chosen when setting up
the network. If DNS is not available then the Networker Server and Hostname can be entered as
below.
It is also possible to select the version to be restored: Instead of a particular version, the backup
may be recovered from a particular point in time.
Once the backup has been successfully setup, you can move to the next stage to setup the
configuration.
Configuration
Before recovery can begin, the machine configuration information created earlier must be loaded into
the recovery environment from the backup. This is performed by selecting the Restore
Configuration From Backup option:
If the location of the configuration information was changed during Recording System Information,
you need to enter the location chosen here.
Selecting the Options item displays any additional options that can be applied at this point.
Create Minimum Size LVs: This option ensures that the logical volumes created are of the
smallest size such that the data to restore fits. This option is useful if you are recovering to a
Only Recover Root VG: By default, all volume groups are recovered. This option is useful in
situations where data is stored on a second data-only volume group which is not included in the
backup.
Keep going on non-root failures: By default all failures are considered fatal and immediately halt
a recovery. If this option is selected only failures that prevent the restoration of volumes and file-
systems directly associated with the root volume group halt a recovery.
Note: the minimum size calculation is performed when the configuration information is
recorded but it is re-calculated when the backup is accessed.
The Edit Drive Mapping item is used to modify the disks that the backup is restored to. In the
example given, a system using two disks is mapped to a use only one disk during recovery. Should
the drive display spread over more than one screen you can navigate forwards Ctrl +N or the previous
page Ctrl +P
When recovering to fewer disks, any volume groups other than rootvg which cannot be re-created are
dropped. However, a volume group spanning more than one physical volume can be restored to a
single volume provided that volume has enough capacity. In the case of mirrored volume groups the
mirroring is split if the mapping indicates this.
The Edit Group Mapping item is used to modify the disks that the backup is restored to by
selecting which disks belong to which volume group. In the example given both disks are given to the
rootvg volume group and the testvg volume group is given one, so is not restored. The same rules for
recovering to fewer disks as were used for Edit Drive Mapping are used here.
Perform Recovery
Once the configuration has been restored, it is possible to start the recovery. When this option is
selected, the backup location is tested and a confirmation dialogue presented:
Note: as soon as the automatic recovery is started, ALL data is destroyed on the disks
being recovered to
Once the recovery is complete, you are presented with a dialogue indicating that the machine can
be rebooted:
Copy Logs
Once the recovery is complete, a menu is opened containing options for viewing and copying the log
files.
The Summary Log Files item is used to present a summary of the warnings, errors and
informational items that occurred during the recovery for immediate inspection.
Copying the log files to an NFS share allows any problems encountered during recovery and
Setup
The setup menu allows you to setup networking (exactly as the first stage of automatic recovery),
setup the backup location (exactly as the second stage of automatic recovery) and mount an NFS
volume.
Configuration
The configuration menu allows you to restore the configuration from backup, select which Volume
Groups should be restored and modify configuration options.
This menu allows you to control precisely which volume groups should be restored. This is in
contrast to the options provided for automatic recovery, which only allow restoration of all volume
groups, or just the root.
The third option allows you to test the backup location for connectivity before performing a restore. It
is recommended that this step is always performed before recovering a system.
Note: running an earlier phase after a later phase, such as running LogicalVolumes after
Restore, will RESET ANY WORK DONE BY THE LATER PHASE. You will therefore have
to run the remaining phases as well to complete the restore
This option may be used to restart a stalled recovery from the next phases to run until completion.
For example, if the recovery stopped at the FileSystems stage, then running from Mounting to
MakeBootable should result in a fully restored system.
Additionally, there is a special option which attempts to work out the last known phase:
9.1.4 Trouble-shooting
Recovery:
If the automatic recovery fails at any point, then it may be possible to continue to recover the system
by continuing the recovery at the next phase.
> Disrec::ERROR "The following commands failed in the last phase run"
> ...
> Disrec::ERROR "Review the logs and correct any errors before proceeding
> Disrec::ERROR "to the next phase (MOUNTING)"
Then it may be possible to get a working system by running the phases from Mounting until the final
phase, MakeBootable.
All phases between Mounting and MakeBootable may be run by selecting Run Between Two
Phases and selecting the Mounting and MakeBootable phases. If preferred, the phases may be run
individually by selecting Run Single Phase.
Once the final phase, MakeBootable, has been run then it is possible to reboot the machine.
However, we recommend copying log files to an accessible location (to an NFS server for example)
before performing the reboot.
Terminal:
The recovery environment uses the terminal 'aixterm' by default. However, for some displays or
hardware this is not always appropriate. If the terminal is unusable, for example if the menu-options
do not correctly line up, then it may be desirable to change the terminal type. This can be performed
by selecting 'Exit to Shell' and running the environment using a different terminal.
Typing 'terms' produces a list of the terminals available. However, is typically quite long, it may be
useful to try one of the following common terminal types first:
xterm
vt102
vt100
lft
For example, typing 'TERM=vt100 dr' restarts the recovery environment using the vt100 terminal
type.
If using Putty as the terminal emulator selecting ISO-8859-1:1998 as the remote character set
can help correct character translation issues:
The Copy Log Files option asks for a location and creates a date-stamped archive of the logs in the
directory given.
>logs-12:54-10092010.tar.gz
Note: it is important that the directory selected is an NFS mount, as all information in the
recovery environment is lost on reboot
Support Hours
Out-of-Hours support available to customers with a valid Support Agreement - Severity 1 issues* only
*Severity 1 issues are defined as: a production server failure, cannot perform recovery or actual loss
of data occurring.
**For details on dates of UK Bank Holidays, please see www.cristie.com/support/
Cristie Software Ltd. are continually expanding their product range in line with the latest
technologies. Please contact the Cristie Sales Office for the latest product range.