This week we had three different account collections which got stuck. We've had this every now and then but three times in a week is different.
We have to restart the acm.
Is there another way to just kill this particular collection or do you really have to restart?
And what could be the reason of these collection problems?
Do you have any further questions or may I mark this discussion thread as answered?
The issue is not completely solved. But we are in contact with RSA.
Anyway we found it was enough to restart the agent so it's possible to get the collectors going without bringing down the GUI.
Yes if the Agent is a remote agent it can be restarted to resolve this issue. For a local Agent acm must be restarted.
Did you identify if any of your collectors matched had settings matching the known issues outlined in the CAUSE statement in the KB I referenced?
I don't think the AC or the AD configuration is the cause. I also happens with collectors which have been running for years.
I recommend you open an RSA Custom Support case for this issue.
The engineer can then conduct an evidence based analysis specific to the particular issue you are having and provide guidance.
The most common collection problems are these two issues with the AD LDAP collector as mentioned in the cause statement. Referrals should always be disabled for AD. We recommend you do not use DNS round robin for AD.
000037395 - Collection is stuck in the data collection phase in RSA Identity Governance & Lifecycle 7.x
If the issue is with another type of collector then there are other possibilities but not easy solutions.
This issue occurs if the collector does not fail, but also does return a result. The Agent does not have any way to determine that the collector should be abandoned.
A proposal to create collector timeout settings is available on RSA Ideas. Please vote for it.
Option to define collection duration (time out)