IBM HTTP Server policies
This section provides details of the supplied IBM HTTP Server policies to enable you to customize them for your specific environment.
If you simply want to get started using the policy as quickly as possible (with no customization), see Getting started with policies. In order to gain a better understanding of policies, see Policy architecture.
There is no Minimal level agent policy for IBM HTTP Server monitoring. Basic is the lowest available level.
The Basic level agent policy for IHS references the IHS_POLICY_Basic Object Policy Group. Because IBM HTTP Server monitoring is very simple and contains only one object type (the IHS Server), there are no nested Groups. Instead, IHS_POLICY_Basic directly references the IHS_Server_Basic Object Policy. That Object Policy then applies one Policy Action (IHS_Server_UseHistoryTemplate_SAMPLE) which activates history collection for the key attributes of all IHS Servers using the rules defined in the SAMPLE history template.
The basic design of the IHS_Standard_EventTarget policies is very similar to the Basic level with the agent policy referring to the IHS_POLICY_Standard_EventTarget Object Policy Group which in turn refers to one Object Policy (IHS_Server_Standard_EventTarget). As you would expect, this Object Policy includes the IHS_Server_UseHistoryTemplate_SAMPLE Policy Action but then adds a set of Policy Actions to add event templates (see below).
At the Standard level, three such event templates are associated with each IHS Server. These generate events to the Event Target when the IHS Server is unavailable, when the Request Rate to that Server exceeds a threshold and when the free percentage of the server drops below a threshold. Note that these thresholds are customizable.
The IHS_Agents_ALL_EventTarget agent policy follows the same structure as the Standard level but adds additional Policy Actions. As you can see below, the additional actions add event templates to notify when the Byte Rate into the server exceeds a threshold. This can distinguish the effect of a smaller number of large requests vs simply using the Request Rate as in the Standard level. Additionally, the ALL level adds escalating alerts not just when the percentage of idle servers is low but when the absolute number falls below a threshold. Finally, an alert is added for when too little traffic is coming into the IHS Server (LowRate). Again, the thresholds for each of these events is customizable.