Thanks to all that have replied. We're still busy researching and testing but we did catch a break yesterday and it appears that the problem could be the same that Wol mentioned. We have had reports of CPU spikes in the past, but these were only reported in off hours and usually the following day or later. Yesterday however, they noticed a process that was Spiking using 100% CPU. We were able to track it to the user and find out exactly what they were doing. The user confirmed that they would just exit from the telnet session window (Windows Telnet) and not log off properly. When I tested this, the Process immediately began spiking.
We're thinking and hoping that the slow downs that we have experienced are the cumulative affects of multiple bad disconnects and CPU spikes. I have contacted Rocket to see if/when this issue has been resolved and we will keep monitoring the server. Again thanks for the help and input. We're still going through several of the suggestions to see what impact they may have. Curt Stewart -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Wols Lists Sent: Friday, April 08, 2011 11:54 AM To: [email protected] Subject: Re: [U2] Extreme slow down on Universe periodically On 08/04/11 14:04, Curt Stewart wrote: > Has anyone in the group experienced this issue and found a resolution? > > > Any and all suggestions would be appreciated. > > Environment > Universe 10.2.10 > Windows Server 2003 > Virtual Machine (sorry don't know any details on this component) > 34 user license, only about 15 active > This is an old problem, that might well be resolved on 10.2, but do you have any rogue uvtelnets? We used to get that - they'd get disconnected somehow from their user, and then grab 100% cpu. Kill them, and things went back to normal (we had to crash them, they wouldn't kill, by attaching the debugger or something). Charles mentioned the network. Are your network buffers too large? I'm beginning to pick up from my reading on a hardware/software mismatch at the TCP level, and network slowdown (and catastrophic failure) are the result. Cheers, Wol _______________________________________________ U2-Users mailing list [email protected] http://listserver.u2ug.org/mailman/listinfo/u2-users _______________________________________________ U2-Users mailing list [email protected] http://listserver.u2ug.org/mailman/listinfo/u2-users
