AIOps

AIOps

Join this online group to communicate across IBM product users and experts by sharing advice and best practices with peers and staying up to date regarding product enhancements.

 View Only

Agent Deployment (ITM in Plain English)

By IMWUC Community Team posted Mon February 19, 2018 08:06 AM

  

by John Round

This blog entry presents an overview of the complete agent deployment process.

ITM Agent Depot Location

Referred to as either an agent depot or deployment depot. It is an installation directory on the monitoring server from which you deploy agents and maintenance packages across your environment. The depots are usually located and maintained independently on the hub monitoring server and one or more remote monitoring servers. The default location of the depot is;

Windows:

Linux and UNIX:

To change the default location, edit the KBBENV monitoring server configuration file located in the

When you remotely deploy to a server, the depot you use belongs to the monitoring server to which the operating system agent reports. However, a hub monitoring server and one or more remote monitoring servers can use a single depot. You just need to locate the depot in a central location that is accessible by all of the monitoring servers. See Locating and sizing the remote deployment depot for more information.

NOTE: You cannot create an agent depots on a z/OS monitoring server. There is no command to create a depot. The location of the depot is either a default location or one specified by the DEPOTHOME environment variable in the KBBENV file.

 

Populating the ITM Agent Depot

You can populate the ITM Depot during an installation (see Populating the agent depot from the installation image) or by using the tacmd addBundles command (see Populating the agent depot with the tacmd addBundles command). A bundle is the agent installation image and any prerequisites.

To view available bundles that you can load into a depot from the installation media or image, use tacmd listBundles. To list the types of bundles available in either the local or remote agent depot, use tacmd viewDepot.

NOTE: You do not install agents remotely directly from the installation media. You copy bundles of installation code from the installation media into the depot. You must run tacmd addBundles locally on a monitoring server containing a depot.

Command reference: tacmd addBundles; tacmd listBundles; tacmd viewDepot.

 

Agent Deployment

The command to use to remotely deploy operating system agents is tacmd createNode.You must deploy operating system agents to a target destination before remotely deploying application agents to that server.Operating system agents monitor operating system performance, but also include the required infrastructure for remote deployment and maintenance.

This command uses the following protocols:

Server message block (SMB), use primarily for Windows servers

Secure shell (SSH), only SSH version 2

Remote execution (REXEC)

Remote shell (RSH)

By default, it tries all available protocols until a protocol works. You can also specify which protocol to use.

This command also runs a prerequisite check automatically. To verify that the prerequisite checker runs, issue the tacmd createNode command using the EXECPREREQCHECK=Y and COLLECTALL=Y options. When using these two options, the prerequisite checker writes its results to /opt/IBM/ITM/logs/checkprereq_results. A prerequisite check can also be run locally as a stand-alone check (a tool needs to be copied from the installation media for the appropriate platform to the target computer), or remotely by using the tacmd checkPrereq command.

NOTE: Although you can install an operating system agent remotely, you cannot control it remotely via the command line or via the Tivoli Enterprise Portal client.

The command to use to remotely deploy application agents (non-OS agents) is tacmd addSystem. To use this deployment method, you must first deploy an operating system agent on the target server. You can also use the Tivoli Enterprise Portal client to remotely deploy application agents by right clicking on the server instance and selecting "Add Managed System". And you can control them remotely via the command line or via the Tivoli Enterprise Portal client.

Command reference: tacmd createNode; tacmd checkPrereq; tacmd addSystem.

 

Useful Deployment Commands

See the IBM Tivoli Monitoring Command Reference for the full syntax and description of all the commands.

General

Depot management

Deployment and configuration

Agent management

Miscellaneous

help

viewDepot

describeSystemType

startAgent

viewNode

login

listBundles

addSystem

stopAgent

listSystems

logout

addBundles

configureSystem

restartAgent

configurePortalServer

 

removeBundles

createNodeAgent

viewAgent

createNodeAgent

     

updateAgent

clearDeployStatus


NOTE: To take action, such as create, modify, delete, view, an operating system user must first authenticate against a server by issuing a tacmd login command.

 

Example Deployment Process

This is a summary of steps used to install and manage the Tivoli Log Agent on a remote Linux computer using tacmd commands for deployment.

  1. Login to the Tivoli Enterprise Monitoring Server.

  2. listBundles Show the details of one or more bundles in the CD.

  3. addBundles Add one or more bundles to the local deployment depot.

  4. viewDepot Show the agents that you can install from the depot.

  5. listSystems Show a list of agents, managed systems, or both.

  6. createNode Deploy (install) an OS agent to a target remote computer.

  7. listSystems List the current systems showing the new node that you create.

  8. viewNode Show the characteristics of a node.

  9. addSystem Deploy a monitoring agent and prerequisite software to a node. The computer must already have an OS agent installed.

  10. viewNode Show the characteristics of the node to verify that the agent is installed.

  11. stopAgent Stop the agent.

  12. startAgent Start the agent. Can also use the restartAgent command.

 

Click here for the video demonstrating the above process.

 

Checking Deployment Status

Every deployment command (including checkPrereq) has a transaction ID. This is the long number displayed when a deployment command is successfully executed. You can use this number with the command tacmd getDeployStatus to check the progress of any deployment command. You can also view this information in the Tivoli Enterprise Portal client by right clicking Enterprise in the Navigator Physical view and selecting Workspace > Deployment Status Summary.

Commend reference: tacmd getDeployStatus; clearDeployStatus.

 

Deployment Troubleshooting

You can add options to the tacmd commands to increase the logging in the debug logs.

JLOG_LEVEL=DEBUG_MAX

Prevent the tacmd addSystem command from removing the temporary directory in the target workstation.

AUTOCLEAN=NO

 

Setting RAS1 trace levels

On Windows systems:

On the computer where the Tivoli Enterprise Monitoring Server is installed, select Start > Programs > IBM Tivoli Monitoring > Manage Tivoli Enterprise Monitoring Services.

Right-click the Tivoli Enterprise Monitoring Server service.

Select Advanced > Edit Trace Parms > to display the Trace Parameters window.

Type (UNIT:kdy all) in the Enter RAS1 Filters field.

Accept the defaults for the rest of the fields.

Click OK to set the new trace options.

Click Yes to recycle the service.

 

On Linux systems, set the following variable in $CANDLEHOME/config/lz.ini:

KBB_RAS1=ERROR(UNIT:kdy ALL)(UNIT:kdd ALL)

 

On UNIX systems other than Linux, set the following variable in $CANDLEHOME/config/ux.ini:

KBB_RAS1=ERROR (UNIT:kdy ALL) (UNIT:kdd ALL)

 

For dynamic tracing, please see Dynamically modify trace settings for an IBM Tivoli Monitoring component.

0 comments
10 views

Permalink