Managing message delays

A message delay is the phenomenon that occurs when new, incoming messages that arrive on the management server are propagated with a significant time delay to different internal consumers (for example, WMI, console, status engine).

The biggest contributor to message delays is the OvEpMessageFilter component of the OvEpMessageActionServer service. The OvEpMessageFilter component validates whether messages originate from authorized managed nodes. All messages must pass this validation before they can be routed to consumers. OvEpMessageFilter uses the OvEpNodeCache component which is responsible for determining valid managed nodes.

Message validation is fast for all messages that include an AgentId and where the sending node is set up as an HPOM managed node. In all other cases the validation must pass a multiple level name resolution process which can cause significant delays.

Possible root causes for message delays are:

• Messages from deleted nodes

  Messages are sent from nodes that have been deleted on the management server but the agents have not been uninstalled and are still sending messages.

• Messages from agentless nodes

  A proxy node gathers information (for example, SNMP traps) from agentless nodes and forwards this information as HPOM messages to the management server. The validation of messages from proxy nodes may also take considerable time as these messages do not contain an AgentId.

In the worst case, delivery of new messages stops completely and the message pipeline hangs.

Related Topics:

• Detect nodes that cause a message delay
• Configure message delay detection