
Hi,
I am monitoring (not creating) some IP SLA operations and want some clarification on the polling intervals. Can someone please confirm the operation of the IP SLA polling? I found the following article ( http://knowledgebase.solarwinds.com/kb/questions/1461/How+does+IP+SLA+Manager+poll+operations+when+there+are+multiple+operations+being+polled+at+different+frequencies%3F ) but still have some questions.
IP SLA Manager will divide all monitored operations on a node into groups based on their polling frequency. For any given group of operations, IP SLA Manager will find the operation with the most frequent polling interval in the group and then poll twice that frequency for that group on the node. Consider the following example:
There are four operations being monitored on a single node. These operations have the following polling frequencies:
Operations A and B would be put in the same group because their polling intervals are both <= 3 minutes. IP SLA Manager will poll both of these operations simultaneously every 30 seconds because the most frequent of the two operations is 1 minute, and IP SLA Manager polls twice that frequency.
Operations C and D would be put into another group because their polling intervals are both <= 6 minutes. IP SLA Manager will poll both of these operations simultaneously every 2 minutes because the most frequent of the two operations is 4 minutes, and IP SLA Manager polls twice that frequency.
1) Where do the <= 3 minutes and <= 6 minutes groups come from?
2) Does the polling interval mentioned equate to the operation frequency?
3) Does this mean if one node has 5 IP SLA operations and they all have a frequency of 60 seconds, the node will be polled every 30 seconds for IP SLA operation status/statistics?
4) And finally, I'm unsure if I should create a new topic for this question but, in the Manage IP SLA Operations screen, is it possible to filter nodes or group by a custom property (not just the current: [No Grouping], Status, Type, Source, Target)?
Currently using IP SLA Manager 3.5.1.
Thankyou
Hi,
1. The groups are system-defined, tuned to decrease the overload polling load. They cannot be customized.
2. The example seems to be not precise. When aggregating, first, the operation frequencies are divided by 2 (the polling needs to be twice as fast, that's correct here), these are the polling intervals. And finally, the intervals are grouped together based on the following table:
3. Correct.
4. IPSLA Operations don't have custom properties, so you could just filter by nodes' custom properties, anyway this is not supported. The list contains only basic set of IPSLA operation properties.
Cheers Tom. That cleared up my questions.