Configuring Infrastructure Management self-monitoring

To ensure reliability and performance of the Infrastructure Management server, events and automated email are generated when the number of monitored instances exceed predefined thresholds. The thresholds are determined automatically depending on your environment size (small, medium, or large.) For more information, see Hardware requirements to support small, medium, and large environments. If you make changes to your environment, follow these steps to reconfigure the predefined thresholds.

By default, addition of new PATROL Agents and Integration Services are blocked until the number of monitored instances is reduced below the predefined thresholds. To reduce the number of instances, see Recommendations to reduce the instance count.

Threshold	Event severity	Email generated?	Addition of PATROL Agents and Integration Services	Addition of new instances from PATROL Agents
80% of maximum limit	Minor	Yes	Open	Open
90% of maximum limit	Major	Yes	Open	Open
100% of maximum limit	Critical	Yes	Blocked	Blocked

To configure the blocking of PATROL Agents and Integration Services

On the Infrastructure Management server, access the <installationDirectory>\pw\pronto\conf\pronet.conf file.
Copy the following properties:
pronet.new_is_pa.block=true
pronet.selfmon.mail=true

Paste these properties into the <installationDirectory>\pw\custom\conf\pronet.conf file and modify the values as required.

Property	Description	Values
pronet.new_is_pa.block	Blocks the addition of PATROL Agents and Integration Services when the number of monitored instances reach 100% of the predefined threshold.	true (default) false
pronet.selfmon.mail	Generates automated email for self-monitoring events.	true (default) false

Restart the Infrastructure Management server for the changes to take effect.

To reconfigure the predefined thresholds after a hardware configuration change

If you make changes to your hardware configuration, like upgrading the RAM, CPU, or both, perform the following steps to reconfigure the thresholds.

On the Infrastructure Management server, copy the following properties from the <installationDirectory>\pw\pronto\conf\pronet.conf file.
# Max # of Instances for large setup
pronet.instance.large.max=250000
# Max # of Instances for medium setup
pronet.instance.medium.max=60000
# Max # of Instances for small setup
pronet.instance.small.max=30000
Paste these properties into the <installationDirectory>\pw\custom\conf\pronet.conf file and modify the values as required.
Restart the Infrastructure Management server for the changes to take effect.

Viewing self-monitoring events

When the number of monitored instances exceed predefined thresholds, events are generated and displayed in the TrueSight console.

After you log in to the TrueSight console, go to Monitoring > Events.
When the thresholds for number of monitored instances are exceeded, Minor, Major, and Critical events are generated.
When the number of monitored instances reduce to below the thresholds, Information events are generated and the original events are closed.

Recommendations to maintain optimum performance

When the number of monitored instances exceed predefined thresholds consistently, you can consider the following options to maintain the optimum performance of your environment. For more information on individual hardware requirements, see Hardware requirements.

Current environment	Recommendation	Additional configurations
Small	Upgrade to a Medium environment	Create a back up. For more information, see Archiving or backing up the Infrastructure Management Server. Stop the server node. Add additional RAM capacity. For more information, see Hardware requirements to support small, medium, and large environments. Log into the Infrastructure Management server. Edit the <install_directory>/pw/custom/conf/pronet.conf and set the pronet.deployment.type=medium. Restart the Infrastructure Management server. (On HA) Repeat the steps on the secondary node. (On HA) Verify the status by running the pw ha status command. The status of the primary node must be Active or the status of the secondary node must be Standby.
Medium	Upgrade to a Large environment	Create a back up. For more information, see Archiving or backing up the Infrastructure Management Server. Stop the server node. Add additional RAM capacity and CPUs. For more information, see Hardware requirements to support small, medium, and large environments. Log into the Infrastructure Management server. Edit the <install_directory>/pw/custom/conf/pronet.conf and set the pronet.deployment.type=large. Restart the Infrastructure Management server. (On HA) Repeat the steps on the secondary node. (On HA) Verify the status by running the pw ha status command. The status of the primary node must be Active or the status of the secondary node must be Standby.
Large	Add an Infrastructure Management Server	Reduce the number of monitor instances. For more information, see Recommendations to reduce the instance count.

Recommendations to reduce the instance count

On reaching the maximum limit, the new incoming instances are automatically shelved in the INSTANCE_WAIT_LIST table of the back-end database. The shelved instances are automatically unshelved in the order of arrival and added to the Infrastructure Management server when you reduce the instance count below the maximum limit. For best practices and sizing guidelines on number of monitoring instances, see General sizing guidelines and best practices for Infrastructure Management.

Follow these recommendations to identify and reduce the instance count:

On the Infrastructure Management operator console, go to the Performance Diagnostics page to identify the instance count by Application classes.
Identify the monitor instances that are no longer relevant, such as virtual servers #, Windows Services, and Windows Job Objects.
Follow one or more of these recommendations to reduce the instance count:
- Disable or delete infrastructure policies.
- Do not add new infrastructure policies until the instance count reduces below the maximum limit.
- Migrate data instances to event-only instances by editing policies.
- Delete the instances marked for deletion by running the pw mfdinst delete command. For more information, see pw-mfdinst.
- Move the newly deployed PATROL Agents to a different Infrastructure Management server to reduce the load. You can retrieve the PATROL Agents list by running the pw patrolagent list command. For more information, see pw-patrolagent-list.