We use Tuxedo Version 8.0, 32-bit, Patch Level 331 middleware, which is connected to Kenan Billling. Recently, observed few servers like om_Customer & ck_GRS servers (specific to Kenan API's) gets always HANG specific to few Databases. Would like to know the root cause for this hang out issue. As a work around, currently, we bounce these specific servers as and when it get hang,resulting few message loss.
Request your help on this. And also, currently, low/high water mark (new server spawning) set as 1:30,3:20, planning to change to 1:60,3:10 for a better tuxedo server performance. Need your comments on this.
In general support for Tuxedo in an ISV application is up to the ISV. That being said, without more information such as the contents of your ULOG, it will be nearly impossible for anyone to tell you why a server is hung. Do they hang indefinitely or only for a short period of time? Does the database report that the servers are still connected? With regards to server spawning, again this is something Kenan would be much better to advise you on as they know the application. Without knowing the function of the servers involved, their performance characteristics, or the load on your system, it would be difficult to provide you with guidance on server spawning.
Oracle Tuxedo Chief Architect
Firstly regarding server hang issue, we did monitor the ULOG while is the server is hang, but, could not find much info'n in ULOG as Tuxedo keep adding the new servers of the same as queue length is getting increased but none of teh requests are processed. Refer the below ULOG info. Is there any generic reasons as to why servers gets hang and steps to check and rectify? please advice.
001554.com100.celcom.com.my!ck_GRS.22433.1.0: 05-19-2012: Tuxedo Version 8.0, 32-bit
001554.com100.celcom.com.my!ck_GRS.22433.1.0: LIBTUX_CAT:262: INFO: Standard main starting
001555.com100.celcom.com.my!ck_GRS.22433.1.0: ck_GRS.c@275: CUST SERVER: 6
001555.com100.celcom.com.my!ck_GRS.22433.1.0: ck_GRS: initiated.
001555.com100.celcom.com.my!ck_GRS.22433.1.0: LIBTUX_CAT:476: WARN: Server 11/1939: client process 9698: lost message
001555.com100.celcom.com.my!ck_GRS.22433.1.0: LIBTUX_CAT:477: WARN: SERVICE=M_PFAn MSG_ID=0 REASON=server died
001555.com100.celcom.com.my!restartsrv.22432.1.-2: 05-19-2012: Tuxedo Version 8.0, 32-bit
Secondly on server spawning, yes, we got the answer as how to adjust the respective parameters.