on 11-04-2013 10:34 AM
Dear All,
we installed new PI 7.31 system almost 4 months and already 2 phases of Upgradation project already live, it's working on last 2 months.
Now a days my PRODUCTION SYSTEM is hanging very freequently when one of my scenario is running it is RFC-->PI--->JDBC . when ever this scenario is running with high load my Production system goes hanging and struck the queues and failing many messages with HTTP 401 error.
Could you please suggest me what are the parameters that I need to change? thank you.
Sateesh
Hello Sateesh,
Are you using dual-stack? If yes, maybe you can try increasing these parameters in nwa:
ume.r3.connection.master.poolmaxsize
ume.r3.connection.master.poolmaxwait
For more information, see this document:
http://scn.sap.com/docs/DOC-16238 under section 2.2.2 HTTP 401 - Authentication Failure
See also SAP Note 821026 - AdapterEngine Messaging URL returns HTTP 401 under high load.
The parameter (for PI 7.3) can be found in Configuration -> Security -> Destinations (UMEBackendConnection)
Hope this helps,
Mark
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello Mark,
Thanks for your reply,
I already followed this note month ago and changed valued of parameters to 50 & 60000 ms.
But now a days my system is hanging not exactly related to HTTP 401 under high load. but still few messages were failing with same reason.
Main thing was Hanging the system.
I need some more suggestions from your end.
Regards,
Sateesh
Hi Mark,
My old XI production we used the value 50, since last 5 years. I can't get trouble on the old system with this value.
but in New PI system also I I used to 50 earlier it was 10(month ago). and after that since last one month the scneario have been running fine but in the last week(it is month end) it was huge messages were processing with include other scenarios and System is going to hang.
My Dialog work process also fine I increased to 50 as compare to my old system.
Is their any other configuration is required to stop the system hang?
Thank you.
Sateesh
Hello Sateesh,
Not sure if the following would help:
1. Use XBTL* for large messages, this can be configured in SXMB_ADM -> Configure Filter for Queue Prioritization
2. Use trace level 1 instead of 3 in prod, also found in SXMB_ADM -> Integration Engine Configuration
3. Disable Logging_Sync in prod, also found in SXMB_ADM -> Integration Engine Configuration
4. Increase server memory/java heap
5. Check for dumps in ST22 (look for TIMEOUTS)
Regards,
Mark
User | Count |
---|---|
90 | |
10 | |
10 | |
10 | |
7 | |
7 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.