Skip to Content

General best practices and guidance for configuration

  • PATROL Agent rules cannot be merged in BMC ProactiveNet Central Monitoring Administration.
  • Non-policy managed Knowledge Modules must be included in the preloaded Knowledge Module list manually outside Central Monitoring Administration.
  • Use caution with policies. They can have global impact.
  • Ensure that the load preloaded KMs agent configuration rule is set to true on all agents before adding the agents to BMC ProactiveNet. This is a default setting in version 9.6 agents and must be ensured for all agents.
  • The PATROL Agent version 9.5 default setting for loadOnlyPreloadedKMs is “Yes”.  If you load a Knowledge Module that is not in the preloaded list on a BMC PATROL console, the console shows that the Knowledge Module is disabled.
  • Define configuration for event integration and monitoring for the availability parameters first.
  • Always establish and follow standards for policies including policy names, precedence, and so on.
  • Do not propagate agent disconnected events to Integration for BMC Remedy Service Desk (and thereby generate tickets in BMC IT Service Management) in a development or test environment and/or from a server related to a staging Integration Service node. When the agents are disconnected from these environments and are connecting to a production Integration Service, the disconnected events are expected and normal.
  • Do not enable policies for production agents until tested. The default setting is disabled.
  • Leverage the ability to test policies in development and QA environments then apply to the production environment from the same Central Monitoring Administration instance in environments that have a single, shared Central Monitoring Administration instance. Leverage selection criteria in monitoring policies for this.  See the Architecture-and-scalability-best-practices for details.
  • Set marked for delete in the Central Monitoring Administration of the Central Servers to 30 minutes to deal with the agents that are temporarily in the Central Server. Do not do this for Central Servers that are also used for trending production.
  • Allow large domain collection like VMware vSphere discovery to complete before reviewing it in the BMC ProactiveNet Operations Console. The hierarchy has to be built and the views may look odd until it is complete.
  • The PATROL Agents buffer data and events when they cannot connect to the Integration Service node. The default setting is set to buffer for 30 minutes. Do not increase this setting especially in large environments, as larger volumes of buffered data being processed after a reconnection can potentially cause performance issues.
  • 9.6 PATROL Agents support the use of customer overlay configuration files in the lib directory to manage alterations to the config.default file settings. The customer overlay configuration files are named <name>.current where <name> is whichever file name that you want. Any .current configuration files that exist in the lib directory are read after the config.default file and before the pconfig database is read.
    Note    : Overlay configuration files do not work with /ConfigData rules. These rules should be sent by Central Monitoring Administration only.
  • Be very careful while deleting and/or disabling policies. When you delete or disable a policy, all corresponding monitored instances on all related managed nodes are immediately be marked for deletion. If you have a policy that needs significant editing, and you want to delete that policy, it is best to do the following:
    1. Create a new policy with a lower precedence number that overrides the older policy.
    2. After you have tested, validated, and enabled the new policy in production, delete the old policy.
  • Use the most descriptive names available when defining instances to a monitor. For example, leverage the “Display name” for monitoring Windows Services instead of the “Service name”. This makes it easier to identify the service in the BMC ProactiveNet UI.
  • Use the default setting of the KPI mode for the BMC ProactiveNet Server. Do not disable the KPI mode causing all trended data in the BMC ProactiveNet Server to have baselines calculated for them. This is also recommended for earlier BMC ProactiveNet versions. This is not as significant an issue with BMC ProactiveNet 9.6 if the new BMC ProactiveNet 9.6 capability to query BMC PATROL data from the BMC ProactiveNet UI is leveraged. However, the recommendation must still be observed.
  • The Agent Query user function in Central Monitoring Administration has two scenarios regarding security. These scenarios also involve different communication channels to the agents and are described as follows:
    • Seamless access - If the user has the "Allow trusted connections to PATROL Agents" permission, they can access agents through the agent query function without manually logging into the agent. In this scenario, the communication path to the agents from Central Monitoring Administration is through the Agent Controller to the Integration Service node and on to the agent. The communication leverages the existing connection to the agent for policy management.
    • Agent logon required access – If the user does not have the "Allow trusted connections to PATROL Agents" permission, the Agent Query function requires that the user manually log into the agent. In this scenario, the communication path to the agents from Central Monitoring Administration is to port 3181 on the agent, or the port that the agent is listening on. This scenario bypasses the Integration Service node connection path and must be accounted for in the architecture plans and related firewall setting if it is to be leveraged.

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*

BMC ProactiveNet 9.6