Re: TLB IPI Wait Errors & APIC interrupt handling

Amy Abascal-Turner (amy@vasoftware.com)
Fri, 14 Dec 2001 15:11:03 -0800 (PST)


Oh, I should mention that they're using 2.2.19

--
Amy Abascal-Turner                   510-687-6741
Sr. Mgr. Product Support       amy@vasoftware.com
VA SOFTWARE CORP.       http://www.vasoftware.com
-------------------------------------------------

On Fri, 14 Dec 2001, Amy Abascal-Turner wrote:

> Hi! > > VA Linux (Software) is in dire need of some kernel help. > If you know anything about TLB IPI and/or APIC interrupt > handling, we will gladly pay well for your expertise and > time. We need the help on a very short timeline. > > A brief description of the problem is below. Please email > me or call me at 408-621-7054 (or the number below) if you > can help or would like further details. > > Thanks! > > Amy > -- > Amy Abascal-Turner 510-687-6741 > Sr. Mgr. Product Support amy@vasoftware.com > VA SOFTWARE CORP. http://www.vasoftware.com > ------------------------------------------------- > > Brief History > > A scientific computing cluster of 600 VA 1220s has been > experiencing various problems under heavy loading > conditions under production scenarios. VA engineers have > been dedicated to identifying and solving these problems > and although the situation has vastly improved, it is > still not completely resolved. The primary issue > remaining is random rebooting in SMP mode contributing to > instability as a cluster. > > Technical Problems > > Random Reboots: > "TLB IPI Wait" errors, possibly indicative of kernel > deadlock. This will require kernel-development expertise > to resolve. > > Reboots possibly indicative of APIC interrupt handling > which will require kernel development expertise to > resolve. > > Internal Clock Skew: resolved on by replacing motherboard > on most nodes experiencing problem. Suspect that some > clock problems are side-effects of the APIC/TLB issues > noted. > > Resolution > > In order to resolve the Reboot / TLB/IPI issues, the > expertise of a kernel developer is required. We are > currently identifying resources to contract with to > analyze the problem(s) and implement a solution. > > > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ >

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/