Hi,
Currently, in order to check whether an active agent is still running, I'm monitoring the agent by letting it send a hearbeat every 15 minutes. I've attached a trigger to it so it signals an alarm in case no hearbeat is received within 35 minutes (hence two missed heartbeats).
This works on most of the hosts that I'm monitoring. However, on two hosts it seems that the hearbeat is sometimes not being sent by the agent. On one host I have it multiple times a day that the alarm is (in my opion, incorrectly) raised.
Related to this issue, I have two questions:
Currently, in order to check whether an active agent is still running, I'm monitoring the agent by letting it send a hearbeat every 15 minutes. I've attached a trigger to it so it signals an alarm in case no hearbeat is received within 35 minutes (hence two missed heartbeats).
This works on most of the hosts that I'm monitoring. However, on two hosts it seems that the hearbeat is sometimes not being sent by the agent. On one host I have it multiple times a day that the alarm is (in my opion, incorrectly) raised.
Related to this issue, I have two questions:
- How to debug/found out why some agents are not sending their data (item) in time?
- Is there a possibility to implement the nodata() trigger on host level? So instead of binding it to a specific item, is it possible to fire an alarm in case no data at all for the host is received for some time?