In the PollSumary.log is frequent to see warning messages like this:
yyyy-mm-dd hh:mm:ss,pid | WARN | 300000-thread-1 | PollSummary | dm.core.common.EndOfCycleMessage 500 | 188 - com.ca.im.data-collection-manager.core.impl - 3.2.0.RELEASE-126 | | Number of poll requests triggered does not match the number of polled items: polledItemCount=17471, pollRequestCount=9839, lateDroppedPollRequestCount=0, pollGroupId=4071, statCycleStartTime=1511169300000
Are the warning messages seen in PollSummary.log saying “Number of poll requests triggered does not match the number of polled items:…” a symptom of some problem in getting data from Data Collector or from the monitored devices?
This type of message is frequently seen in PollSummary.log and normally is benign.
Some devices might have slow response that could cause the response to end outside of the poll cycle time; as consequence these warnings are generated.
These messages means that the Data Collector can't reach a device (due to timeout), and the device is 'throttled', meaning that the Data Collector won't poll it for some time because its known to be down / not reachable.
The Data Collector is functioning as designed here, and it’s warning about missed polling because of this reason.
Most important messages we would rather need to pay attention in PollSummary.log are those with "responsesReceivedFromJmsCount=0" (in Data Aggregator PollSummary.log) and "responsesSentToJmsCount=0" (in Data Collector PollSummary.log).
These messages indicate that there is no polling data received from the DC or sent to the DA; and the reason would need to be investigated.