Given that data, wouldn't opening a ticket with Microsoft be worth it? If you don't know what is causing it, how will ripping and replacing it help?
*ASB *(Professional Bio <http://about.me/Andrew.S.Baker/bio>) *Technology Services that Maximize Business Results... * On Thu, Apr 14, 2011 at 10:38 AM, James Rankin <[email protected]>wrote: > I wouldn't normally ask this, but I'm kind of at my wits' end. Is there > anyone out there who could maybe have a look at a Process Monitor logfile > and tell me if they can see any kind of "smoking gun" in there? > > The situation is this - we have 2008 R2 Terminal Servers that occasionally > will start treading water, resulting in horrendous logon times for users. > We've tried disabling just about everything, GPOs, AppSense, EdgeSight, > SCOM, antivirus, we have patched them to the hilt with every hotfix we can > find for every piece of software, run countless monitors and logs, sent > details to various support teams, even had a Citrix consultant on site to > offer his opinion, yet still the problem exists. We have carefully monitored > the apps in use on the problem systems (which seem to be completely random) > and can find no correlation between application usage and the occurrence of > this issue. The servers have been monitored by several different tools, > native and otherwise, and do not red-line in any way apart from occasional > spikes of memory usage and page faults, but nothing happens that seems to > justify the terrible performance slowdown that occurs. The servers are > physical Compaq DL360 G6 systems with 16GB of RAM and 16 CPUs. > > Luckily I managed to capture a ProcMon log the last time this problem > happened (usually running it causes the server to come to a complete halt, > more or less). Rather interestingly, when the logon completed, the ProcMon > log was actually running two minutes behind "real-time" - it took two > minutes to catch up with what was actually happening "live" on the server! > I've had a good hunt through this, but I'm more used to looking for > application issues than trying to troubleshoot a logon with ProcMon, and I > simply don't know what to look for to try and identify the causes of the > slowdown. Microsoft's removal of the user environment debug logging in 2008 > and up is a real pain, as it was (fairly) straightforward to troubleshoot > the logon process previously. > > I am fairly sure that the problem is something intrinsic to the system - > i.e. not caused by a third-party piece of software. I'm on the verge of > recommending that the whole server farm is ripped and replaced but I want to > make sure I've covered all my bases before I go down that route. > > If anyone can help with this, please ping me offline and I'll gladly > provide access to the (monstrously large, given that the logon I was > monitoring took six minutes) log file. Or if anyone has any pointers that > they think might help with the performance, I'll also gladly take them on > board. > > TIA, > > > > JRR > > ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~ --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to [email protected] with the body: unsubscribe ntsysadmin
