Done all of that. But with the latest outage we have it narrowed down to the restarting of the mid-tiers that is causing the issues. Sent to BMC on one of the restarts a multi gig log file.
The strange item is that we refreshed our QA and staging system, with production data and the issue does not show up there. Howard -----Original Message----- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Longwing, LJ CTR MDA/IC Sent: Friday, February 01, 2013 12:49 PM To: arslist@ARSLIST.ORG Subject: Re: Slow system restart/arsystem errors after restarting the mid-tiers on 7.6.4 Howard, I highly recommend turning on API/SQL logging to the same file, and then startup the Mid-Tier. At the end of the 'unavailable' period you should have a VERY good log file that will be able to be analyzed by arwklga to give you the reasons for your errors. It will tell you what API calls were taking the longest as well as the SQL in question, from that you should be able to analyze what in your system is slow and in turn tune it. -----Original Message----- From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Howard Richter Sent: Friday, February 01, 2013 10:40 AM To: arslist@ARSLIST.ORG Subject: Slow system restart/arsystem errors after restarting the mid-tiers on 7.6.4 ** Happy Friday all, We have been seeing a very strange error since our go-live with 7.6.4 (and working with no luck with BMC support), when we restart our mid-tiers. We are on windows server 2008 and using MSSQL 2008. What we have seen is we will shut down our two customer facing arservers (as well as our integration arserver), as well as our three mid-tiers. Then we bring up our three arservers (integration first, then the two customer facing ones), with no issues. After testing them with auto creation of an incident via email (an integration we created) and some poking around using the old user tool, we bring up one of our mid-tiers (i.e. start up the tomcat engine). At that point, the arserver that the mid-tier starts to talk to, starts to show database time out errors and you can no longer log in via the user tool. It takes about 30 minutes for the system (the arserver/mid-tier) to stabilize and then we are able to bring up the other two mid-tiers. We have not seen this issue in our QA system (which is all on VM) and our staging system. They both have somewhat the same architecture and we have verified the mid-tier configuration/patch releases. Since we are still working with BMC support (for over 3 months on this), I just wanted to see if anyone had seen this type of issue on other 7.6.4 systems. Take care, Howard _ARSlist: "Where the Answers Are" and have been for 20 years_ ____________________________________________________________________________ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years" _______________________________________________________________________________ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
