Thanks Mark, I'll look into that config option, and try again with top open.
In this case I was doing a home backup to a 1TB WD Caviar black formatted as ext3. I also have a raid0 with 2 other non-WD sata drives, and a single WD velociraptor I can test with. It also doesn't sound too far off that FF could be the culprit (mentioned above), as I have it open all the time, and so far it's been the first place I've noticed these hiccups. That could be coincidence though, as I've pretty much always got it open and these hiccups are system wide. On Mon, Aug 23, 2010 at 10:03 AM, Mark Knecht <markkne...@gmail.com> wrote: > On Sun, Aug 22, 2010 at 12:01 PM, Alan Warren <bluemoonsh...@gmail.com> > wrote: > > Hello, > > > > I am having some system performance issues with this kernel release. I > have > > a SMP machine (dual xeon nehalem 8 core / 16 threads) with 24gb non-ecc > > memory. > > > > On occasion (seems random so far) my system feels like a Pentium II > trying > > to cope with Vista. For example, I was in the middle of tar'ing a > semi-large > > file and noticed all of my apps came to a crawl. Scrolling in firefox, > > typing in the terminal, or trying to navigate in my file manager resulted > in > > breif "pauses" that came in waves. On one occasion my system froze > > completely and I had to manually reset the machine. (that was with > > 2.6.35-r1) > > > > I didn't activate anything "new" in this kernel release that I don't > > normally activate. ie, no cpuidle driver > > > > Is there a proper venu for debugging such matters, or should I just wait > for > > this kernel to go prime-time? > > > > Thanks for your time, > > Alan > > > > Hi Alan, > Sorry for the problems. I've seen them also in the recent past. In > my case it was on new hardware so I couldn't say it was due to a > specific kernel release. > > 1) What happens when you watch top while doing the tar? Do you by any > chance see large wait times in top? (Hit '1' to watch all CPUs) If so > the problem could well be how the kernel is dealing with writing data > back to the hard drive. I had this problem with the WD Green drives. > When I changed to WD RAID Edition drives (1/2 the storage for 30% more > money) the problems disappeared. > > 2) If it's not the drive issue then there is a kernel option called (I > think) RCU_CPU_STALL_DETECTION (or something like that. If you turn it > on I may generate a trace of what's keeping a core busy to long. > Mileage will vary. > > Good luck, > Mark > >
