Zenoss Core Installation Guide r5.1.5 d1051.16.207

Download as pdf or txt
Download as pdf or txt
You are on page 1of 32

Zenoss Core Installation

Guide
Release 5.1.5

Zenoss, Inc.

www.zenoss.com
Zenoss Core Installation Guide

Copyright © 2016 Zenoss, Inc. All rights reserved.

Zenoss and the Zenoss logo are trademarks or registered trademarks of Zenoss, Inc., in the United States and other countries. All other trademarks, logos,
and service marks are the property of Zenoss or other third parties. Use of these marks is prohibited without the express written consent of Zenoss, Inc., or
the third-party owner.

Amazon Web Services, AWS, and EC2 are trademarks of Amazon.com, Inc. or its affiliates in the United States and/or other countries.

Flash is a registered trademark of Adobe Systems Incorporated.

Oracle, the Oracle logo, Java, and MySQL are registered trademarks of the Oracle Corporation and/or its affiliates.

Linux is a registered trademark of Linus Torvalds.

RabbitMQ is a trademark of VMware, Inc.

SNMP Informant is a trademark of Garth K. Williams (Informant Systems, Inc.).

Sybase is a registered trademark of Sybase, Inc.

Tomcat is a trademark of the Apache Software Foundation.

VMware is a registered trademark or trademark of VMware, Inc. in the United States and/or other jurisdictions.

Windows is a registered trademark of Microsoft Corporation in the United States and other countries.

All other companies and products mentioned are trademarks and property of their respective owners.

Part Number: 1051.16.207

Zenoss, Inc.
11305 Four Points Drive
Bldg 1 - Suite 300
Austin, Texas 78726

2
Contents
About this guide.................................................................................................................. 4

Chapter 1: Installing on hosts with internet access........................................6


Installing a master host.................................................................................................................................................... 6
Installing resource pool hosts.......................................................................................................................................18
ZooKeeper ensemble configuration............................................................................................................................ 26
Adding hosts to the default resource pool................................................................................................................ 31
Deploying Zenoss Core................................................................................................................................................. 32

3
Zenoss Core Installation Guide

About this guide


Zenoss Core Installation Guide provides detailed procedures for installing Zenoss Core.

Note Zenoss strongly recommends reviewing the Zenoss Core Planning Guide carefully before using this guide.

Related publications
Title Description 
Zenoss Core Administration Guide Provides an overview of Zenoss Core architecture and
features, as well as procedures and examples to help use
the system.
Zenoss Core Configuration Guide Provides required and optional configuration procedures
for Zenoss Core, to prepare your deployment for
monitoring in your environment.
Zenoss Core Installation Guide Provides detailed information and procedures for
creating deployments of Control Center and Zenoss
Core.
Zenoss Core Planning Guide Provides both general and specific information for
preparing to deploy Zenoss Core.
Zenoss Core Release Notes Describes known issues, fixed issues, and late-breaking
information not already provided in the published
documentation set.
Zenoss Core Upgrade Guide Provides detailed information and procedures for
upgrading deployments of Zenoss Core.

Additional information and comments

Zenoss welcomes your comments and suggestions regarding our documentation. To share your comments, please
send an email to [email protected]. In the email, include the document title and part number. The part number
appears at the end of the list of trademarks, at the front of this guide.

Change history

The following list associates document part numbers and the important changes to this guide since the previous
release. Some of the changes involve features or content, but others do not. For information about new or changed
features, refer to the Zenoss Core Release Notes.
1051.16.207
Update release numbers.
1051.16.176
Update release numbers.
1051.16.153
Update release numbers.
1051.16.146
Update release numbers.
1051.16.125
Refine the procedure for creating the application data thin pool.

4
About this guide

1051.16.118
Add support for Zenoss Core 5.1.2.
Add a substep to create the docker override directory.
1051.16.111
Add this document change history.
Add chapters describing how to install the Zenoss Core appliance.
Chapters are organized into parts.
Docker configuration steps now add the storage driver flag (-s devicemapper) to the /etc/
sysconfig/docker file.
Docker needs a longer startup timeout value, to work around a known Docker issue with the devicemapper
driver. Docker configuration steps now include adding TimeoutSec=300.
Rather than editing /lib/systemd/system/docker.service, Docker configuration steps now
include adding a systemd override file.
Add a symlink to /tmp in /var/lib/docker.
Update the commands for starting and testing a ZooKeeper ensemble.
Add a procedure for updating the SERVICED_ZK value on resource pool hosts that are not members of a
ZooKeeper ensemble.
Add a reference topic for the ZooKeeper variables required on hosts in a Control Center cluster.
Add step to install the Nmap Ncat package, which is used to check ZooKeeper ensemble status.
1051.16.060
Planning information is now in the Zenoss Core Planning Guide.
Information about how to start and configure Zenoss Core is now in the Zenoss Core Configuration Guide.

5
Zenoss Core Installation Guide

Installing on hosts with internet access 1


The procedures in this chapter install Control Center and Zenoss Core on one or more Red Hat Enterprise Linux
(RHEL) 7.1 or 7.2 hosts, or one or more CentOS 7.1 or 7.2 hosts. To use the procedures in this chapter, all Control
Center cluster hosts must have internet access.

You may create a single-host or a multi-host deployment. For production use, Zenoss strongly recommends creating
a multi-host deployment that includes a minimum of three real or virtual machines. For more information about
deploying Control Center and Zenoss Core, refer to the Zenoss Core Planning Guide.

Note For optimal results, review this chapter thoroughly before starting the installation process.

Installing a master host


Perform the procedures in this section to install Control Center and Zenoss Core on a master host.

Verifying candidate host resources


This procedure determines whether a hosts's hardware resources and operating system are sufficient to serve as a
Control Center master host.

1 Log in to the candidate host as root, or as a user with superuser privileges.


2 Verify that the host implements the 64-bit version of the x86 instruction set.

uname -m

■If the output is x86_64, the architecture is 64-bit. Proceed to the next step
■If the output is i386/i486/i586/i686, the architecture is 32-bit. Stop this procedure and select a
different host.
3 Verify that name resolution works on this host.

hostname -i
If the result is not a valid IPv4 adddress, add an entry for the host to the network nameserver, or to /etc/
hosts.
4 Verify that the host's numeric identifier is unique.
Each host in a Control Center cluster must have a unique host identifier.

hostid

6
Installing on hosts with internet access

5 Determine whether the available, unused storage is sufficient.


a Display the available storage devices.

lsblk --output=NAME,SIZE
b Compare the available storage with the amount required for a Control Center master host.
For more information, refer to the Zenoss Core Planning Guide.
6 Determine whether the available memory and swap is sufficient.
a Display the available memory.

free -h
b Compare the available memory with the amount required for a master host in your deployment.
For more information, refer to the Zenoss Core Planning Guide.
7 Update the operating system, if necessary.
a Determine which release is installed.

cat /etc/redhat-release
If the result includes 7.0, perform the following substeps.
b Update the operating system.

yum update -y
c Restart the system.

reboot

Preparing storage for the master host


In addition to the storage required for its operating system, a Control Center master host requires the following
storage areas:

■ A local partition for Docker data, configured as a device mapper thin pool.
■ A local partition for Control Center internal services data, formatted with the XFS file system.

Note Control Center internal services include ZooKeeper, which requires consistently fast storage. Zenoss
recommends using a separate, high-performance storage resource for Control Center internal services. For
example, a drive that is configured with only one primary partition, which eliminates contention by other
services.
■ A local or remote primary partition for Zenoss Core data, configured as a device mapper thin pool.
■ A local primary partition, a remote primary partition, or a remote file server, for backups of Zenoss Core data.
The local or remote primary partition is formatted with the XFS file system. A remote file server must provide a
file system that is compatible with XFS.

Note If you are using a primary partition on a local device for backups, ensure that the primary partition for
Control Center internal services data is not on the same device.

For storage sizing information, refer to the Zenoss Core Planning Guide.

For device mapper thin pools, no formatting is required—simply create primary partitons, which are configured in
subsequent procedures. For more information, refer to the Zenoss Core Planning Guide.

7
Zenoss Core Installation Guide

To create the required storage, perform the following procedures.

Note Data present on the primary partitions you select are destroyed in these procedure. Please ensure that data
is backed up elsewhere, or no longer needed, before proceeding.

Creating a file system for internal services

This procedure creates an XFS file system on a primary partition. For more information about primary partitions,
refer to the Zenoss Core Planning Guide.

Note Control Center internal services include ZooKeeper, which requires consistently fast storage. Zenoss
recommends using a separate, high-performance storage resource for Control Center internal services. For example,
a drive that is configured with only one primary partition, which eliminates contention by other services.

1 Log in to the target host as root, or as a user with superuser privileges.


2 Identify the target primary partition for the file system to create.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
For more information about the output of the lsblk command, and about creating primary partitions, refer to
the Zenoss Core Planning Guide.
3 Create an XFS file system.
Replace Partition with the path of the target primary partition:

mkfs -t xfs Partition


4 Add an entry to the /etc/fstab file.
Replace Partition with the path of the primary partition used in the previous step:

echo "Partition \
/opt/serviced/var/isvcs xfs defaults 0 0" >> /etc/fstab
5 Create the mount point for internal services data.

mkdir -p /opt/serviced/var/isvcs
6 Mount the file system, and then verify it mounted correctly.

mount -a && mount | grep isvcs


Example result:

/dev/xvdb1 on /opt/serviced/var/isvcs type xfs


(rw,relatime,seclabel,attr2,inode64,noquota)

Creating a file system for backups


To perform this procedure, you need a host with at least one unused primary partition, or a remote file server.

The Control Center master host requires local or remote storage space for backups of Control Center data. This
procedure includes steps to create an XFS file system on a primary partition, if necessary, and steps to mount a file
system for backups. For more information about primary partitions, refer to the Zenoss Core Planning Guide.

Note If you are using a primary partition on a local device for backups, ensure that the primary partition for
Control Center internal services data is not on the same device.

8
Installing on hosts with internet access

1 Log in to the target host as root, or as a user with superuser privileges.


2 Optional: Identify the target primary partition for the file system to create, if necessary.
Skip this step if you are using a remote file server.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
For more information about the output of the lsblk command, and about creating primary partitions, refer to
the Zenoss Core Planning Guide.
3 Optional: Create an XFS file system, if necessary.
Skip this step if you are using a remote file server.
Replace Partition with the path of the target primary partition:

mkfs -t xfs Partition


4 Create an entry in the /etc/fstab file.

Replace File-System-Specification with one of the following values:

■ the path of the primary partition used in the previous step


■ the remote server specification

echo "File-System-Specification \
/opt/serviced/var/backups xfs defaults 0 0" >> /etc/fstab
5 Create the mount point for backup data.

mkdir -p /opt/serviced/var/backups
6 Mount the file system, and then verify it mounted correctly.

mount -a && mount | grep backups


Example result:

/dev/sdb3 on /opt/serviced/var/backups type xfs


(rw,relatime,seclabel,attr2,inode64,noquota)

Preparing the master host operating system


This procedure prepares a RHEL/CentOS 7.1 or 7.2 host as a Control Center master host.

1 Log in to the candidate master host as root, or as a user with superuser privileges.
2 Add an entry to /etc/hosts for localhost, if necessary.
a Determine whether 127.0.0.1 is mapped to localhost.

grep 127.0.0.1 /etc/hosts | grep localhost


If the preceding commands return no result, perform the following substep.
b Add an entry to /etc/hosts for localhost.

echo "127.0.0.1 localhost" >> /etc/hosts


3 Disable the firewall, if necessary.
This step is required for installation but not for deployment. For more information, refer to the Zenoss Core
Planning Guide.

9
Zenoss Core Installation Guide

a Determine whether the firewalld service is enabled.

systemctl status firewalld.service

■ If the result includes Active: inactive (dead), the service is disabled. Proceed to the next
step.
■ If the result includes Active: active (running), the service is enabled. Perform the following
substep.
b Disable the firewalld service.

systemctl stop firewalld && systemctl disable firewalld


On success, the preceding commands display messages similar to the following example:

rm '/etc/systemd/system/dbus-org.fedoraproject.FirewallD1.service'
rm '/etc/systemd/system/basic.target.wants/firewalld.service'
4 Optional: Enable persistent storage for log files, if desired.
By default, RHEL/CentOS systems store log data only in memory or in a small ring-buffer in the /run/log/
journal directory. By performing this step, log data persists and can be saved indefinitely, if you implement
log file rotation practices. For more information, refer to your operating system documentation.

mkdir -p /var/log/journal && systemctl restart systemd-journald


5 Disable Security-Enhanced Linux (SELinux), if installed.
a Determine whether SELinux is installed.

test -f /etc/selinux/config && grep '^SELINUX=' /etc/selinux/config

If the preceding commands return a result, SELinux is installed.


b Set the operating mode to disabled.
Open /etc/selinux/config in a text editor, and change the value of the SELINUX variable to
disabled.
c Confirm the new setting.

grep '^SELINUX=' /etc/selinux/config


6 Enable and start the Dnsmasq package.

systemctl enable dnsmasq && systemctl start dnsmasq


7 Install the Nmap Ncat utility.
The utility is used to verify ZooKeeper ensemble configurations. If you are installing a single-host
deployment, skip this step.

yum install -y nmap-ncat


8 Install and configure the NTP package.
a Install the package.

yum install -y ntp

10
Installing on hosts with internet access

b Set the system time.

ntpd -gq
c Enable the ntpd daemon.

systemctl enable ntpd


d Configure ntpd to start when the system starts.
Currently, an unresolved issue associated with NTP prevents ntpd from restarting correctly after a reboot.
The following commands provide a workaround to ensure that it does.

echo "systemctl start ntpd" >> /etc/rc.d/rc.local


chmod +x /etc/rc.d/rc.local
9 Install the Zenoss repository package.
a Install the package.

rpm -ivh http://get.zenoss.io/yum/zenoss-repo-1-1.x86_64.rpm


b Clean out the yum cache directory.

yum clean all


10 Reboot the host.

reboot

Installing Docker and Control Center


This procedure installs and configures Docker, and installs Control Center.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Install Docker 1.9.0, and then disable accidental upgrades.
a Add the Docker repository to the host's repository list.

cat > /etc/yum.repos.d/docker.repo <<-EOF


[dockerrepo]
name=Docker Repository
baseurl=https://yum.dockerproject.org/repo/main/centos/7
enabled=1
gpgcheck=1
gpgkey=https://yum.dockerproject.org/gpg
EOF
b Install Docker 1.9.0.

yum clean all && yum makecache fast


yum install -y docker-engine-1.9.0
c Open /etc/yum.repos.d/docker.repo with a text editor.
d Change the value of the enabled key from 1 to 0.
e Save the file and close the text editor.
3 Create a symbolic link for the Docker temporary directory.

11
Zenoss Core Installation Guide

Docker uses its temporary directory to spool images. The default directory is /var/lib/docker/tmp. The
following command specifies the same directory that Control Center uses, /tmp. You can specify any directory
that has a minimum of 10GB of unused space.
a Create the docker directory in /var/lib.

mkdir /var/lib/docker
b Create the link to /tmp.

ln -s /tmp /var/lib/docker/tmp
4 Create a systemd override file for the Docker service definition.
a Create the override directory.

mkdir -p /etc/systemd/system/docker.service.d
b Create the override file.

cat <<EOF > /etc/systemd/system/docker.service.d/docker.conf


[Service]
TimeoutSec=300
EnvironmentFile=-/etc/sysconfig/docker
ExecStart=
ExecStart=/usr/bin/docker daemon \$OPTIONS -H fd://
EOF
c Reload the systemd manager configuration.

systemctl daemon-reload
5 Install Control Center.
Control Center includes a utility that simplifies the process of creating a device mapper thin pool.

yum clean all && yum makecache fast


yum --enablerepo=zenoss-stable install -y serviced-1.1.7
6 Create a device mapper thin pool for Docker data.
a Identify the primary partition for the thin pool to create.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
b Create the thin pool.

Replace Path-To-Device with the path of an unused primary partition:

serviced-storage create-thin-pool docker Path-To-Device


On success, the result includes the name of the thin pool, which always starts with /dev/mapper.
7 Configure and start the Docker service.
a Create variables for adding arguments to the Docker configuration file.
The --exec-opt argument is a workaround for a Docker issue on RHEL/CentOS 7.x systems.

Replace Thin-Pool-Device with the name of the thin pool device created in the previous step:

myDriver="-s devicemapper"
myFix="--exec-opt native.cgroupdriver=cgroupfs"

12
Installing on hosts with internet access

myFlag="--storage-opt dm.thinpooldev"
myPool="Thin-Pool-Device"
b Add the arguments to the Docker configuration file.

echo 'OPTIONS="'$myDriver $myFix $myFlag'='$myPool'"' \


>> /etc/sysconfig/docker
c Start or restart Docker.

systemctl restart docker


The initial startup takes up to a minute, and may fail. If the startup fails, repeat the previous command.
8 Configure name resolution in containers.
Each time it starts, docker selects an IPv4 subnet for its virtual Ethernet bridge. The selection can change; this
step ensures consistency.
a Identify the IPv4 subnet and netmask docker has selected for its virtual Ethernet bridge.

ip addr show docker0 | grep inet


b Open /etc/sysconfig/docker in a text editor.
c Add the following flags to the end of the OPTIONS declaration.

Replace Bridge-Subnet with the IPv4 subnet docker selected for its virtual bridge, and replace Bridge-Netmask
with the netmask docker selected:

--dns=Bridge-Subnet --bip=Bridge-Subnet/Bridge-Netmask

For example, if the bridge subnet and netmask is 172.17.0.1/16, the flags to add are --dns=172.17.0.1
--bip=172.17.0.1/16.

Note Leave a blank space after the end of the thin pool device name, and make sure the double quote
character (") is at the end of the line.

d Restart the Docker service.

systemctl restart docker

Installing Zenoss Core


This procedure installs Zenoss Core and configures the NFS server.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Install Zenoss Core.

yum --enablerepo=zenoss-stable install -y zenoss-core-service


3 Configure and restart the NFS server.
Currently, an unresolved issue prevents the NFS server from starting correctly. The following commands provide a
workaround to ensure that it does.
a Open /lib/systemd/system/nfs-server.service with a text editor.
b Change rpcbind.target to rpcbind.service on the following line:

Requires= network.target proc-fs-nfsd.mount rpcbind.target

13
Zenoss Core Installation Guide

c Reload the systemd manager configuration.

systemctl daemon-reload

Configuring Control Center


This procedure creates a thin pool for application data and customizes key configuration variables of Control
Center.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Configure Control Center to serve as the master and as an agent.

The following variables configure serviced to serve as both master and agent:
SERVICED_AGENT
Default: 0 (false)
Determines whether a serviced instance performs agent tasks. Agents run application services
scheduled for the resource pool to which they belong. The serviced instance configured as the master
runs the scheduler. A serviced instance may be configured as agent and master, or just agent, or just
master.
SERVICED_MASTER
Default: 0 (false)
Determines whether a serviced instance performs master tasks. The master runs the application
services scheduler and other internal services, including the server for the Control Center browser interface.
A serviced instance may be configured as agent and master, or just agent, or just master. Only one
serviced instance in a Control Center cluster may be the master.
a Open /etc/default/serviced in a text editor.
b Find the SERVICED_AGENT declaration, and then change the value from 0 to 1.
The following example shows the line to change:

# SERVICED_AGENT=0
c Remove the number sign character (#) from the beginning of the line.
d Find the SERVICED_MASTER declaration, and then change the value from 0 to 1.
The following example shows the line to change:

# SERVICED_MASTER=0
e Remove the number sign character (#) from the beginning of the line.
f Save the file, and then close the editor.
3 Create a thin pool for Zenoss Core data.
a Identify the primary partition for the thin pool to create, and the amount of space available on the primary
partition.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
For more information about the output of the lsblk command and primary partitions, refer to the Zenoss
Core Planning Guide.
b Create a variable for 50% of the space available on the primary partition for the thin pool to create.
The thin pool stores application data and snapshots of the data. You can add storage to the pool at any time.

14
Installing on hosts with internet access

Replace Half-Of-Available-Space with 50% of the space available in the primary partition, in gigabytes. Include
the symbol for gigabytes (G) after the numeric value.

myFifty=Half-Of-Available-SpaceG
c Create the thin pool.

Replace Path-To-Device with the path of the target primary partition:

serviced-storage create-thin-pool -o dm.basesize=$myFifty \


serviced Path-To-Device
On success, the result includes the name of the thin pool, which always starts with /dev/mapper.
4 Configure Control Center with the name of the thin pool for Zenoss Core data.

The Control Center configuration file is /etc/default/serviced. (For more information about
serviced configuration options, refer to the Control Center online help.)
a Open /etc/default/serviced in a text editor.
b Locate the SERVICED_FS_TYPE declaration.
c Remove the number sign character (#) from the beginning of the line.
d Add SERVICED_DM_THINPOOLDEV immediately after SERVICED_FS_TYPE.

Replace Thin-Pool-Name with the name of the thin pool created previously:

SERVICED_DM_THINPOOLDEV=Thin-Pool-Name
e Save the file, and then close the editor.
5 Optional: Specify an alternate private subnet for Control Center, if necessary.

The default private subnet may already be in use in your environment. The following variable configures
serviced to use an alternate subnet:
SERVICED_VIRTUAL_ADDRESS_SUBNET
Default: 10.3
The 16-bit private subnet to use for serviced's virtual IPv4 addresses. RFC 1918 restricts private
networks to the 10.0/24, 172.16/20, and 192.168/16 address spaces. However, serviced accepts any
valid, 16-bit, IPv4 address space for its private network.
a Open /etc/default/serviced in a text editor.
b Locate the SERVICED_VIRTUAL_ADDRESS_SUBNET declaration, and then change the value.
The following example shows the line to change:

# SERVICED_VIRTUAL_ADDRESS_SUBNET=10.3
c Remove the number sign character (#) from the beginning of the line.
d Save the file, and then close the editor.

User access control


Control Center provides a browser interface and a command-line interface.

To gain access to the Control Center browser interface, users must have login accounts on the Control Center
master host. (Pluggable Authentication Modules (PAM) is supported.) In addition, users must be members of the
Control Center administrative group, which by default is the system group, wheel. To enhance security, you may
change the administrative group from wheel to any non-system group.

15
Zenoss Core Installation Guide

To use the Control Center command-line interface, users must have login accounts on the Control Center master
host, and be members of the docker user group. Members of the wheel group, including root, are members of
the docker group.

Adding users to the default administrative group


This procedure adds users to the default administrative group of Control Center, wheel. Performing this
procedure enables users with superuser privileges to gain access to the Control Center browser interface.

Note Perform this procedure or the next procedure, but not both.

1 Log in to the host as root, or as a user with superuser privileges.


2 Add users to the system group, wheel.

Replace User with the name of a login account on the master host.

usermod -aG wheel User

Repeat the preceding command for each user to add.

Note For information about using Pluggable Authentication Modules (PAM), refer to your operating system
documentation.

Configuring a regular group as the Control Center administrative group


This procedure changes the default administrative group of Control Center from wheel to a non-system group.

Note Perform this procedure or the previous procedure, but not both.

1 Log in to the Control Center master host as root, or as a user with superuser privileges.
2 Create a variable for the group to designate as the administrative group.
In this example, the name of group to create is serviced. You may choose any name or use an existing group.

GROUP=serviced
3 Create a new group, if necessary.

groupadd $GROUP
4 Add one or more existing users to the new administrative group.

Replace User with the name of a login account on the host:

usermod -aG $GROUP User

Repeat the preceding command for each user to add.


5 Specify the new administrative group in the serviced configuration file.
The following variable specifies the administrative group:
SERVICED_ADMIN_GROUP
Default: wheel
The name of the Linux group on the Control Center master host whose members are authorized to use
the Control Center browser interface. You may replace the default group with a group that does not have
superuser privileges.
a Open /etc/default/serviced in a text editor.

16
Installing on hosts with internet access

b Find the SERVICED_ADMIN_GROUP declaration, and then change the value from wheel to the name
of the group you chose earlier.
The following example shows the line to change:

# SERVICED_ADMIN_GROUP=wheel
c Remove the number sign character (#) from the beginning of the line.
d Save the file, and then close the editor.
6 Optional: Prevent root users and members of the wheel group from gaining access to the Control Center
browser interface, if desired.
The following variable controls privileged logins:
SERVICED_ALLOW_ROOT_LOGIN
Default: 1 (true)
Determines whether root, or members of the wheel group, may gain access to the Control Center
browser interface.
a Open /etc/default/serviced in a text editor.
b Find the SERVICED_ALLOW_ROOT_LOGIN declaration, and then change the value from 1 to 0.
The following example shows the line to change:

# SERVICED_ALLOW_ROOT_LOGIN=1
c Remove the number sign character (#) from the beginning of the line.
d Save the file, and then close the editor.

Enabling use of the command-line interface


This procedure enables users to perform administrative tasks with the Control Center command-line interface by
adding individual users to the docker group.

1 Log in to the Control Center master host as root, or as a user with superuser privileges.
2 Add users to the Docker group, docker.

Replace User with the name of a login account on the host.

usermod -aG docker User

Repeat the preceding command for each user to add.

Starting Control Center


This procedure starts the Control Center service, serviced.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Start serviced.

systemctl start serviced

To monitor progress, enter the following command:

journalctl -flu serviced -o cat

The serviced daemon invokes docker to pull its internal services images from Docker Hub. The Control
Center browser and command-line interfaces are unavailable until the images are installed and the services are
started. The process takes approximately 5-10 minutes. When the message Trying to discover my
pool repeats, Control Center is ready for the next steps.

17
Zenoss Core Installation Guide

3 Note Perform this step only if you are installing a single-host deployment.
Optional: Add the master host to the default resource pool.

Replace Hostname-Or-IP with the hostname or IP address of the Control Center master host:

serviced host add Hostname-Or-IP:4979 default

If you enter a hostname, all hosts in your Control Center cluster must be able to resolve the name, either through
an entry in /etc/hosts, or through a nameserver on your network.

Isolating the master host in a separate resource pool

Note If you are configuring a single-host deployment, skip this procedure.

Control Center enables or just performs rapid recovery from application service failures. When Control Center
internal services and application services share a host, application failures can limit recovery options. Zenoss
strongly recommends isolating the Control Center master host in a separate resource pool.

This procedure creates a new resource pool for the Control Center master host, and then adds the master host to
the pool.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Create a new resource pool named master.

serviced pool add master


3 Add the master host to the master resource pool.

Replace Hostname-Or-IP with the hostname or IP address of the Control Center master host:

serviced host add Hostname-Or-IP:4979 master

If you enter a hostname, all hosts in your Control Center cluster must be able to resolve the name, either through
an entry in /etc/hosts, or through a nameserver on your network.

Installing resource pool hosts


Note If you are installing a single-host deployment, skip this section.

Control Center resource pool hosts run the application services scheduled for the resource pool to which they
belong, and for which they have sufficent RAM and CPU resources.

Zenoss Core has two broad categories of application services: Infrastructure and collection. The services associated
with each category can run in the same resource pool, or can run in separate resource pools.

For improved reliability, two resource pool hosts are configured as nodes in an Apache ZooKeeper ensemble. The
storage required for ensemble hosts is slightly different than the storage required for all other resource pool hosts:
Each ensemble host requires a separate primary partition for Control Center internal services data, in addition to the
primary partition for Docker data. Unless the ZooKeeper service on the Control Center master host fails, their roles
in the ZooKeeper ensemble do not affect their roles as Control Center resource pool hosts.

Note The hosts for the ZooKeeper ensemble require static IP addresses, because ZooKeeper does not support
hostnames in its configurations.

18
Installing on hosts with internet access

Repeat the procedures in the following sections for each host you wish to add to your Control Center deployment.

Verifying candidate host resources


This procedure determines whether a hosts's hardware resources and operating system are sufficient to serve as a
Control Center resource pool host.
Perform this procedure on each resource pool host in your deployment.

1 Log in to the candidate host as root, or as a user with superuser privileges.


2 Verify that the host implements the 64-bit version of the x86 instruction set.

uname -m

■ If the output is x86_64, the architecture is 64-bit. Proceed to the next step
■ If the output is i386/i486/i586/i686, the architecture is 32-bit. Stop this procedure and select a
different host.
3 Verify that name resolution works on this host.

hostname -i
If the result is not a valid IPv4 adddress, add an entry for the host to the network nameserver, or to /etc/
hosts.
4 Verify that the host's numeric identifier is unique.
Each host in a Control Center cluster must have a unique host identifier.

hostid
5 Determine whether the available, unused storage is sufficient.
a Display the available storage devices.

lsblk --output=NAME,SIZE
b Compare the available storage with the amount required for a resource pool host in your deployment.
In particular, resource pool hosts that are configured as nodes in a ZooKeeper ensemble require an
additional primary partition for Control Center internal services data.
For more information, refer to the Zenoss Core Planning Guide.
6 Determine whether the available memory and swap is sufficient.
a Display the available memory.

free -h
b Compare the available memory with the amount required for a resource pool host in your deployment.
For more information, refer to the Zenoss Core Planning Guide.
7 Update the operating system, if necessary.
a Determine which release is installed.

cat /etc/redhat-release
If the result includes 7.0, perform the following substeps.
b Update the operating system.

yum update -y

19
Zenoss Core Installation Guide

c Restart the system.

reboot

Preparing a resource pool host


This procedure prepares a RHEL/CentOS 7.1 or 7.2 host as a Control Center resource pool host.

1 Log in to the candidate resource pool host as root, or as a user with superuser privileges.
2 Add an entry to /etc/hosts for localhost, if necessary.
a Determine whether 127.0.0.1 is mapped to localhost.

grep 127.0.0.1 /etc/hosts | grep localhost


If the preceding commands return no result, perform the following substep.
b Add an entry to /etc/hosts for localhost.

echo "127.0.0.1 localhost" >> /etc/hosts


3 Disable the firewall, if necessary.
This step is required for installation but not for deployment. For more information, refer to the Zenoss Core
Planning Guide.
a Determine whether the firewalld service is enabled.

systemctl status firewalld.service

■ If the result includes Active: inactive (dead), the service is disabled. Proceed to the next
step.
■ If the result includes Active: active (running), the service is enabled. Perform the following
substep.
b Disable the firewalld service.

systemctl stop firewalld && systemctl disable firewalld


On success, the preceding commands display messages similar to the following example:

rm '/etc/systemd/system/dbus-org.fedoraproject.FirewallD1.service'
rm '/etc/systemd/system/basic.target.wants/firewalld.service'
4 Optional: Enable persistent storage for log files, if desired.
By default, RHEL/CentOS systems store log data only in memory or in a small ring-buffer in the /run/log/
journal directory. By performing this step, log data persists and can be saved indefinitely, if you implement
log file rotation practices. For more information, refer to your operating system documentation.

mkdir -p /var/log/journal && systemctl restart systemd-journald


5 Disable Security-Enhanced Linux (SELinux), if installed.
a Determine whether SELinux is installed.

test -f /etc/selinux/config && grep '^SELINUX=' /etc/selinux/config

If the preceding commands return a result, SELinux is installed.


b Set the operating mode to disabled.

20
Installing on hosts with internet access

Open /etc/selinux/config in a text editor, and change the value of the SELINUX variable to
disabled.
c Confirm the new setting.

grep '^SELINUX=' /etc/selinux/config


6 Enable and start the Dnsmasq package.

systemctl enable dnsmasq && systemctl start dnsmasq


7 Install and configure the NTP package.
a Install the package.

yum install -y ntp


b Set the system time.

ntpd -gq
c Enable the ntpd daemon.

systemctl enable ntpd


d Configure ntpd to start when the system starts.
Currently, an unresolved issue associated with NTP prevents ntpd from restarting correctly after a reboot.
The following commands provide a workaround to ensure that it does.

echo "systemctl start ntpd" >> /etc/rc.d/rc.local


chmod +x /etc/rc.d/rc.local
8 Install the Nmap Ncat utility.
The utility is used to verify ZooKeeper ensemble configurations. Perform this step only on the two resource
pool hosts that are designated for use in the ZooKeeper ensemble.

yum install -y nmap-ncat


9 Install the Zenoss repository package.
a Install the package.

rpm -ivh http://get.zenoss.io/yum/zenoss-repo-1-1.x86_64.rpm


b Clean out the yum cache directory.

yum clean all


10 Reboot the host.

reboot

Creating a file system for Control Center internal services


This procedure creates an XFS file system on a primary partition.

21
Zenoss Core Installation Guide

Note Perform this procedure only on the two resource pool hosts that are designated for use in the
ZooKeeper ensemble. No other resource pool hosts run Control Center internal services, so no other pool hosts
need a partition for internal services data.

1 Log in to the target host as root, or as a user with superuser privileges.


2 Identify the target primary partition for the file system to create.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
3 Create an XFS file system.
Replace Isvcs-Partition with the path of the target primary partition:

mkfs -t xfs Isvcs-Partition


4 Create the mount point for Control Center internal services data.

mkdir -p /opt/serviced/var/isvcs
5 Add an entry to the /etc/fstab file.
Replace Isvcs-Partition with the path of the primary partition used in the previous step:

echo "Isvcs-Partition \
/opt/serviced/var/isvcs xfs defaults 0 0" >> /etc/fstab
6 Mount the file system, and then verify it mounted correctly.

mount -a && mount | grep isvcs


Example result:

/dev/xvdb1 on /opt/serviced/var/isvcs type xfs


(rw,relatime,seclabel,attr2,inode64,noquota)

Installing Docker and Control Center


This procedure installs and configures Docker, and installs Control Center.

1 Log in to the resource pool host as root, or as a user with superuser privileges.
2 Install Docker 1.9.0, and then disable accidental upgrades.
a Add the Docker repository to the host's repository list.

cat > /etc/yum.repos.d/docker.repo <<-EOF


[dockerrepo]
name=Docker Repository
baseurl=https://yum.dockerproject.org/repo/main/centos/7
enabled=1
gpgcheck=1
gpgkey=https://yum.dockerproject.org/gpg
EOF
b Install Docker 1.9.0.

yum clean all && yum makecache fast


yum install -y docker-engine-1.9.0
c Open /etc/yum.repos.d/docker.repo with a text editor.

22
Installing on hosts with internet access

d Change the value of the enabled key from 1 to 0.


e Save the file and close the text editor.
3 Create a symbolic link for the Docker temporary directory.
Docker uses its temporary directory to spool images. The default directory is /var/lib/docker/tmp. The
following command specifies the same directory that Control Center uses, /tmp. You can specify any directory
that has a minimum of 10GB of unused space.
a Create the docker directory in /var/lib.

mkdir /var/lib/docker
b Create the link to /tmp.

ln -s /tmp /var/lib/docker/tmp
4 Create a systemd override file for the Docker service definition.
a Create the override directory.

mkdir -p /etc/systemd/system/docker.service.d
b Create the override file.

cat <<EOF > /etc/systemd/system/docker.service.d/docker.conf


[Service]
TimeoutSec=300
EnvironmentFile=-/etc/sysconfig/docker
ExecStart=
ExecStart=/usr/bin/docker daemon \$OPTIONS -H fd://
EOF
c Reload the systemd manager configuration.

systemctl daemon-reload
5 Install Control Center.
Control Center includes a utility that simplifies the process of creating a device mapper thin pool.

yum clean all && yum makecache fast


yum --enablerepo=zenoss-stable install -y serviced-1.1.7
6 Create a device mapper thin pool for Docker data.
a Identify the primary partition for the thin pool to create.

lsblk --output=NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT
b Create the thin pool.

Replace Path-To-Device with the path of an unused primary partition:

serviced-storage create-thin-pool docker Path-To-Device


On success, the result includes the name of the thin pool, which always starts with /dev/mapper.
7 Configure and start the Docker service.
a Create variables for adding arguments to the Docker configuration file.
The --exec-opt argument is a workaround for a Docker issue on RHEL/CentOS 7.x systems.

23
Zenoss Core Installation Guide

Replace Thin-Pool-Device with the name of the thin pool device created in the previous step:

myDriver="-s devicemapper"
myFix="--exec-opt native.cgroupdriver=cgroupfs"
myFlag="--storage-opt dm.thinpooldev"
myPool="Thin-Pool-Device"
b Add the arguments to the Docker configuration file.

echo 'OPTIONS="'$myDriver $myFix $myFlag'='$myPool'"' \


>> /etc/sysconfig/docker
c Start or restart Docker.

systemctl restart docker


The initial startup takes up to a minute, and may fail. If the startup fails, repeat the previous command.
8 Configure name resolution in containers.
Each time it starts, docker selects an IPv4 subnet for its virtual Ethernet bridge. The selection can change; this
step ensures consistency.
a Identify the IPv4 subnet and netmask docker has selected for its virtual Ethernet bridge.

ip addr show docker0 | grep inet


b Open /etc/sysconfig/docker in a text editor.
c Add the following flags to the end of the OPTIONS declaration.

Replace Bridge-Subnet with the IPv4 subnet docker selected for its virtual bridge, and replace Bridge-Netmask
with the netmask docker selected:

--dns=Bridge-Subnet --bip=Bridge-Subnet/Bridge-Netmask

For example, if the bridge subnet and netmask is 172.17.0.1/16, the flags to add are --dns=172.17.0.1
--bip=172.17.0.1/16.

Note Leave a blank space after the end of the thin pool device name, and make sure the double quote
character (") is at the end of the line.

d Restart the Docker service.

systemctl restart docker

Configuring and starting Control Center


This procedure customizes key configuration variables of Control Center.

1 Log in to the resource pool host as root, or as a user with superuser privileges.
2 Configure Control Center as an agent of the master host.

The following variable configures serviced to serve as agent:


SERVICED_AGENT
Default: 0 (false)
Determines whether a serviced instance performs agent tasks. Agents run application services
scheduled for the resource pool to which they belong. The serviced instance configured as the master

24
Installing on hosts with internet access

runs the scheduler. A serviced instance may be configured as agent and master, or just agent, or just
master.
SERVICED_MASTER
Default: 0 (false)
Determines whether a serviced instance performs master tasks. The master runs the application
services scheduler and other internal services, including the server for the Control Center browser interface.
A serviced instance may be configured as agent and master, or just agent, or just master. Only one
serviced instance in a Control Center cluster may be the master.

In addition, the following lines need to be edited, to replace {{SERVICED_MASTER_IP}} with the IP
address of the master host:

# SERVICED_ZK={{SERVICED_MASTER_IP}}:2181
# SERVICED_DOCKER_REGISTRY={{SERVICED_MASTER_IP}}:5000
# SERVICED_ENDPOINT={{SERVICED_MASTER_IP}}:4979
# SERVICED_LOG_ADDRESS={{SERVICED_MASTER_IP}}:5042
# SERVICED_LOGSTASH_ES={{SERVICED_MASTER_IP}}:9100
# SERVICED_STATS_PORT={{SERVICED_MASTER_IP}}:8443
a Open /etc/default/serviced in a text editor.
b Find the SERVICED_AGENT declaration, and then change the value from 0 to 1.
The following example shows the line to change:

# SERVICED_AGENT=0
c Remove the number sign character (#) from the beginning of the line.
d Find the SERVICED_MASTER declaration, and then remove the number sign character (#) from the
beginning of the line.
e Globally replace {{SERVICED_MASTER_IP}} with the IP address of the master host.

Note Remove the number sign character (#) from the beginning of each variable declaration that includes
the master IP address.

f Save the file, and then close the editor.


3 Optional: Specify an alternate private subnet for Control Center, if necessary.

The default private subnet may already be in use in your environment. The following variable configures
serviced to use an alternate subnet:
SERVICED_VIRTUAL_ADDRESS_SUBNET
Default: 10.3
The 16-bit private subnet to use for serviced's virtual IPv4 addresses. RFC 1918 restricts private
networks to the 10.0/24, 172.16/20, and 192.168/16 address spaces. However, serviced accepts any
valid, 16-bit, IPv4 address space for its private network.
a Open /etc/default/serviced in a text editor.
b Locate the SERVICED_VIRTUAL_ADDRESS_SUBNET declaration, and then change the value.
The following example shows the line to change:

# SERVICED_VIRTUAL_ADDRESS_SUBNET=10.3
c Remove the number sign character (#) from the beginning of the line.
d Save the file, and then close the editor.

25
Zenoss Core Installation Guide

4 Start the Control Center service (serviced).

systemctl start serviced

To monitor progress, enter the following command:

journalctl -flu serviced -o cat

To install additional resource pool hosts, return to Verifying candidate host resources on page 19.

ZooKeeper ensemble configuration


Note If you are installing a single-host deployment, or if your deployment includes fewer than two resource pool
hosts, skip this section.

Control Center relies on Apache ZooKeeper to coordinate its services. The procedures in this section create a
ZooKeeper ensemble of 3 nodes. To perform these procedures, you need a Control Center master host and a
minimum of two resource pool hosts. Each resource pool host requires a separate primary partition for Control
Center internal services, and each should have a static IP address. For more information about storage requirements,
refer to the Zenoss Core Planning Guide.

Note Zenoss strongly recommends configuring a ZooKeeper ensemble for all production deployments.

A ZooKeeper ensemble requires a minimum of 3 nodes, and 3 nodes is sufficient for most deployments. A 5-node
configuration improves failover protection during maintenance windows. Ensembles larger than 5 nodes are not
necessary. An odd number of nodes is recommended, and an even number of nodes is strongly discouraged.

Note The Control Center ZooKeeper service requires consistently fast storage. Ideally, the primary partition for
Control Center internal services is on a separate, high-performance device that has only one primary partition.

Control Center variables for ZooKeeper


This tables in this section associates the ZooKeeper-related Control Center variables to set in /etc/default/
serviced with the roles that hosts play in a Control Center cluster.

Table 1: Control Center master host

SERVICED_ISVCS_ZOOKEEPER_ID

The unique identifier of a ZooKeeper ensemble node.


Value: 1
SERVICED_ISVCS_ZOOKEEPER_QUORUM

The ZooKeeper node ID, IP address, peer communications port, and leader communications port of each host
in an ensemble. Each quorum definition must be unique, so the IP address of the "current" host is 0.0.0.0.
Value: ZooKeeper-ID@IP-Address:2888:3888,...
SERVICED_ZK

The list of endpoints in the Control Center ZooKeeper ensemble, separated by the comma character (,). Each
endpoint includes the IP address of the ensemble node, and the port that Control Center uses to communicate
with it.
Value: IP-Address:2181,...

26
Installing on hosts with internet access

Table 2: Control Center resource pool host and ZooKeeper ensemble node

SERVICED_ISVCS_ZOOKEEPER_ID

The unique identifier of a ZooKeeper ensemble node.


Value: 2 or 3
SERVICED_ISVCS_ZOOKEEPER_QUORUM

The ZooKeeper node ID, IP address, peer communications port, and leader communications port of each host
in an ensemble. Each quorum definition must be unique, so the IP address of the "current" host is 0.0.0.0.
Value: ZooKeeper-ID@IP-Address:2888:3888,...
SERVICED_ISVCS_START

The list of Control Center internal services to start and run on hosts other than the master host.
Value: zookeeper
SERVICED_ZK

The list of endpoints in the Control Center ZooKeeper ensemble, separated by the comma character (,). Each
endpoint includes the IP address of the ensemble node, and the port that Control Center uses to communicate
with it.
Value: IP-Address:2181,...

Table 3: Control Center resource pool host only

SERVICED_ZK

The list of endpoints in the Control Center ZooKeeper ensemble, separated by the comma character (,). Each
endpoint includes the IP address of the ensemble node, and the port that Control Center uses to communicate
with it.
Value: IP-Address:2181,...

Configuring the master host as a ZooKeeper node


This procedure configures the Control Center master host as a member of the ZooKeeper ensemble.

Note For accuracy, this procedure constructs Control Center configuration variables in the shell and appends
them to /etc/default/serviced. The last step is to move the variables from the end of the file to more
appropriate locations.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Create a variable for each Control Center host to include in the ZooKeeper ensemble.
The variables are used in subsequent steps.

Note Define the variables identically on the master host and on each resource pool host.

Replace Master-Host-IP with the IP address of the Control Center master host, and replace Pool-Host-A-IP and
Pool-Host-B-IP with the IP addresses of the Control Center resource pool hosts to include in the ensemble:

node1=Master-Host-IP
node2=Pool-Host-A-IP
node3=Pool-Host-B-IP

27
Zenoss Core Installation Guide

Note ZooKeeper requires IP addresses for ensemble configuration.

3 Set the ZooKeeper node ID to 1.

echo "SERVICED_ISVCS_ZOOKEEPER_ID=1" >> /etc/default/serviced


4 Specify the nodes in the ZooKeeper ensemble.
You may copy the following text and paste it in your console:

echo "SERVICED_ZK=${node1}:2181,${node2}:2181,${node3}:2181" \
>> /etc/default/serviced
5 Specify the nodes in the ZooKeeper quorum.
ZooKeeper requires a unique quorum definition for each node in its ensemble. To achieve this, replace the IP
address of the current node with 0.0.0.0.
You may copy the following of text and paste it in your console:

q1="[email protected]:2888:3888"
q2="2@${node2}:2888:3888"
q3="3@${node3}:2888:3888"
echo "SERVICED_ISVCS_ZOOKEEPER_QUORUM=${q1},${q2},${q3}" \
>> /etc/default/serviced
6 Clean up the Control Center configuration file.
a Open /etc/default/serviced with a text editor.
b Navigate to the end of the file, and cut the line that contains the SERVICED_ZK variable declaration at that
location.
The value of this declaration specifies 3 hosts.
c Locate the SERVICED_ZK variable near the beginning of the file, and then delete the line it is on.
The value of this declaration is just the master host.
d Paste the SERVICED_ZK variable declaration from the end of the file in the location of the just-deleted
declaration.
e Navigate to the end of the file, and cut the line that contains the SERVICED_ISVCS_ZOOKEEPER_ID
variable declaration at that location.
f Locate the SERVICED_ISVCS_ZOOKEEPER_ID variable near the end of the file, and then delete the line
it is on.
This declaration is commented out.
g Paste the SERVICED_ISVCS_ZOOKEEPER_ID variable declaration from the end of the file in the
location of the just-deleted declaration.
h Navigate to the end of the file, and cut the line that contains the
SERVICED_ISVCS_ZOOKEEPER_QUORUM variable declaration at that location.
i Locate the SERVICED_ISVCS_ZOOKEEPER_QUORUM variable near the end of the file, and then delete
the line it is on.
This declaration is commented out.
j Paste the SERVICED_ISVCS_ZOOKEEPER_QUORUM variable declaration from the end of the file in
the location of the just-deleted declaration.
k Save the file, and then close the text editor.
7 Verify the ZooKeeper environment variables.

egrep '^[^#]*SERVICED' /etc/default/serviced | egrep '(_ZOO|_ZK)'

28
Installing on hosts with internet access

Configuring a resource pool host as a ZooKeeper node


To perform this procedure, you need a resource pool host with an XFS file system on a separate partition, created
previously.
This procedure configures a ZooKeeper ensemble on a resource pool host. Repeat this procedure on each Control
Center resource pool host to add to the ZooKeeper ensemble.

1 Log in to the resource pool host as root, or as a user with superuser privileges.
2 Create a variable for each Control Center host to include in the ZooKeeper ensemble.
The variables are used in subsequent steps.

Note Define the variables identically on the master host and on each resource pool host.

Replace Master-Host-IP with the IP address of the Control Center master host, and replace Pool-Host-A-IP and
Pool-Host-B-IP with the IP addresses of the Control Center resource pool hosts to include in the ensemble:

node1=Master-Host-IP
node2=Pool-Host-A-IP
node3=Pool-Host-B-IP

Note ZooKeeper requires IP addresses for ensemble configuration.

3 Set the ID of this node in the ZooKeeper ensemble.

For Pool-Host-A-IP (node2), use the following command:

echo "SERVICED_ISVCS_ZOOKEEPER_ID=2" >> /etc/default/serviced

For Pool-Host-B-IP (node3), use the following command:

echo "SERVICED_ISVCS_ZOOKEEPER_ID=3" >> /etc/default/serviced


4 Specify the nodes in the ZooKeeper ensemble.
You may copy the following text and paste it in your console:

echo "SERVICED_ZK=${node1}:2181,${node2}:2181,${node3}:2181" \
>> /etc/default/serviced
5 Specify the nodes in the ZooKeeper quorum.
ZooKeeper requires a unique quorum definition for each node in its ensemble. To achieve this, replace the IP
address of the current node with 0.0.0.0.

For Pool-Host-A-IP (node2), use the following commands:

q1="1@${node1}:2888:3888"
q2="[email protected]:2888:3888"
q3="3@${node3}:2888:3888"
echo "SERVICED_ISVCS_ZOOKEEPER_QUORUM=${q1},${q2},${q3}" \
>> /etc/default/serviced

For Pool-Host-B-IP (node3), use the following commands:

q1="1@${node1}:2888:3888"
q2="2@${node2}:2888:3888"

29
Zenoss Core Installation Guide

q3="[email protected]:2888:3888"
echo "SERVICED_ISVCS_ZOOKEEPER_QUORUM=${q1},${q2},${q3}" \
>> /etc/default/serviced
6 Set the SERVICED_ISVCS_START variable, and clean up the Control Center configuration file.
a Open /etc/default/serviced with a text editor.
b Locate the SERVICED_ISVCS_START variable, and then delete all but zookeeper from its list of
values.
c Remove the number sign character (#) from the beginning of the line.
d Navigate to the end of the file, and cut the line that contains the SERVICED_ZK variable declaration at that
location.
The value of this declaration specifies 3 hosts.
e Locate the SERVICED_ZK variable near the beginning of the file, and then delete the line it is on.
The value of this declaration is just the master host.
f Paste the SERVICED_ZK variable declaration from the end of the file in the location of the just-deleted
declaration.
g Navigate to the end of the file, and cut the line that contains the SERVICED_ISVCS_ZOOKEEPER_ID
variable declaration at that location.
h Locate the SERVICED_ISVCS_ZOOKEEPER_ID variable near the end of the file, and then delete the line
it is on.
This declaration is commented out.
i Paste the SERVICED_ISVCS_ZOOKEEPER_ID variable declaration from the end of the file in the
location of the just-deleted declaration.
j Navigate to the end of the file, and cut the line that contains the
SERVICED_ISVCS_ZOOKEEPER_QUORUM variable declaration at that location.
k Locate the SERVICED_ISVCS_ZOOKEEPER_QUORUM variable near the end of the file, and then delete
the line it is on.
This declaration is commented out.
l Paste the SERVICED_ISVCS_ZOOKEEPER_QUORUM variable declaration from the end of the file in
the location of the just-deleted declaration.
m Save the file, and then close the text editor.
7 Verify the ZooKeeper environment variables.

egrep '^[^#]*SERVICED' /etc/default/serviced \


| egrep '(_ZOO|_ZK|_STA)'
8 Pull the required Control Center ZooKeeper image from the master host.
a Identify the image to pull.

serviced version | grep IsvcsImages


Example result:

IsvcsImages: [zenoss/serviced-isvcs:v40 zenoss/isvcs-zookeeper:v3]


b Pull the Control Center ZooKeeper image.

Replace Isvcs-ZK-Image with the name and version number of the ZooKeeper image from the previous
substep:

docker pull Isvcs-ZK-Image

30
Installing on hosts with internet access

Starting a ZooKeeper ensemble


This procedure starts a ZooKeeper ensemble.
The window of time for starting a ZooKeeper ensemble is relatively short. The goal of this procedure is to restart
Control Center on each ensemble node at about the same time, so that each node can participate in electing the
leader.

1 Log in to the Control Center master host as root, or as a user with superuser privileges.
2 In a separate window, log in to the second node of the ZooKeeper ensemble (Pool-Host-A-IP).
3 In another separate window, log in to the third node of the ZooKeeper ensemble (Pool-Host-B-IP).
4 On all ensemble hosts, stop and start serviced.

systemctl stop serviced && systemctl start serviced


5 On the master host, check the status of the ZooKeeper ensemble.

{ echo stats; sleep 1; } | nc localhost 2181 | grep Mode


{ echo stats; sleep 1; } | nc Pool-Host-A-IP 2181 | grep Mode
{ echo stats; sleep 1; } | nc Pool-Host-B-IP 2181 | grep Mode

If nc is not available, you can use telnet with interactive ZooKeeper commands.
6 Optional: Log in to the Control Center browser interface, and then start Zenoss Core and related applications, if
desired.
The next procedure requires stopping Zenoss Core.

Updating resource pool hosts


The default configuration of resource pool hosts sets the value of the SERVICED_ZK variable to the master host
only. This procedure updates the setting to include the full ZooKeeper ensemble.
Perform this procedure on each resource pool host in your Control Center cluster.

1 Log in to the resource pool host as root, or as a user with superuser privileges.
2 Update the variable.
a Open /etc/default/serviced in a text editor.
b Locate the SERVICED_ZK declaration, and then replace its value with the same value used in the
ZooKeeper ensemble nodes.
c Save the file, and then close the editor.
3 Restart Control Center.

systemctl restart serviced

Adding hosts to the default resource pool


Note If you are installing a single-host deployment, skip this section.

This procedure adds one or more resource pool hosts to the default resource pool.

1 Log in to the Control Center master host as root, or as a user with superuser privileges.
2 Add a resource pool host.

31
Zenoss Core Installation Guide

Replace Hostname-Or-IP with the hostname or IP address of the resource pool host to add:

serviced host add Hostname-Or-IP:4979 default

If you enter a hostname, all hosts in your Control Center cluster must be able to resolve the name, either through
an entry in /etc/hosts, or through a nameserver on your network.
3 Repeat the preceding command for each resource pool host in your Control Center cluster.

Deploying Zenoss Core


This procedure adds the Zenoss Core application to the list of applications that Control Center manages, and pulls
application images from Docker Hub.

1 Log in to the master host as root, or as a user with superuser privileges.


2 Add the Zenoss.core application to Control Center.

myPath=/opt/serviced/templates
serviced template add $myPath/zenoss-core-*.json
On success, the serviced command returns the template identifier.
3 Deploy the application.

Replace Template-ID with the template identifier returned in the previous step, and replace Deployment-ID with a
name for this deployment (for example, Dev or Test):

serviced template deploy Template-ID default Deployment-ID

Control Center pulls Zenoss Core images into the local registry. To monitor progress, enter the following
command:

journalctl -flu serviced -o cat

Control Center and Zenoss Core are now installed, and Zenoss Core is ready to be configured for your
environment. For more information, refer to the Zenoss Core Configuration Guide.

32

You might also like