We are in process of updating TES 5.3.1 to 6.2.1 (of course many hiccups and flaws). We have four environments. Three were upgraded to 6.2.1 but not with latest pack. Last environment is the production environment that is scheduled to go Ist week of January 2015.
We were running production on physical cluster windows 2003 and unfortunately last week (both nodes) died in 3 days in apart that forced us to migrate TES 5.3.1 data to VM non cluster 2008 server. We updated TES 5.3.1 with master 407/ client 391 from master 307/client 303.
Today, we noticed the performance issue in this environment where me and my co-worker ( at different time) tried to insert job. scheduler did not insert job in the job activity successfully. Master log did not create any log for insert that particular record. But client did it only saying " insert jobA done". But did not say your request of insert taken successfully. It happened twice and restarting services of master cleared this error. However I did not like the interruption on the production in the middle of the day as it may delay process as I zero down queue and paused job launching. I onntacted Cisco and recommendation is to increased the java memory on master services. So, placed that parameter in the regedit key of tidal services, and that will effect as I will restart the service next time according to Cisco.
My question is to community, asking if anyone seeing this error running on 5.3.1 407 master/client 391 ? Is vm windows 2008 with 16 G memory good?
Can I take more risk running this version further until TES 6.2.1 fixed completely by Cisco?