Dear SDN Community,
I have an enquiry regarding the POWL setup and technical function within SAP HCM. As with many in the community, we are also having an issue with the “Query <NAME> is already open in another session” message when trying to access a POWL in a short space of time in our Performance Management (OSA-POWL-ESS & OSA-POWL-MSS) iView, and have the option to activate POWL shadowing in so that users can run one query twice:
Question 1:
We have an issue with the cached POWL queries not releasing effectively via the portal (Performance Management iViews). When running POWL_D01, cached queries stay there, but they do not prevent the user from accessing the POWL’s again after logging off properly – they are released and cached, but not deleted. Only when a query is not “released” properly, they create this issue. How do I ensure that they effectively release the session? Is this a portal session (DSM) issue or a backend issue?
Question 2:
Are cached POWL queries constrained by the ICM timeout (rdisp/plugin_auto_logout)? For example, if the timeout is set to 30 minutes and POWL shadowing is enabled, can I be guaranteed that both the original as well as the shadow POWL queries will be released after 30 minutes? Or do I have to run POWL_D01 and/or POWL_D07 to delete these? I do not wish for queries to “build up” in the application server and therefore affect performance severely (since we now have two of the same OSA-MSS/ESS query open).
Question 3:
Are the created/cached queries supposed to delete themselves automatically after a period of time? Currently, our queries are “stuck” there as cached queries, even long after the timeout period. Is there any way to mass delete these queries (OSA-POWL-ESS & OSA-POWL-MSS) on a daily basis via a POWL report? Do you recommend taking this approach?
Additional system details:
SAP HR- 604 (SP56)
SAP Basis/ABA/PI_BASIS - 701 (SP12)
Looking forward to your reply and possible technical recommendations in solving this issue. We have logged an OSS message previously, but only workarounds were suggested.
Regards,
Johan