Date: Thu, 12 Nov 2009 09:38:50 -0800 From: Jack Vogel <jfvogel@gmail.com> To: Mike Tancsa <mike@sentex.net> Cc: pyunyh@gmail.com, jfv@freebsd.org, Mykola Dzham <freebsd@levsha.org.ua>, current@freebsd.org Subject: Re: page fault in igb driver on 8.0-RC2 Message-ID: <2a41acea0911120938g60de3ff0o6fbca1c06bdeb7a3@mail.gmail.com> In-Reply-To: <200911121529.nACFT8ad037531@lava.sentex.ca> References: <20091017222314.GB19204@michelle.cdnetworks.com> <200911100021.nAA0LvMG014534@lava.sentex.ca> <2a41acea0911091633u2899b47ewb514b6276ba2cf62@mail.gmail.com> <200911100227.nAA2RtDY015177@lava.sentex.ca> <2a41acea0911101057q4f913e30m9319e52bbd254cfe@mail.gmail.com> <2a41acea0911101120o39fd695cpa325736d00b11587@mail.gmail.com> <200911102018.nAAKI3s7021614@lava.sentex.ca> <20091111203101.GC15449@michelle.cdnetworks.com> <2a41acea0911111419v5ae5682cy12613e1bd9a9c325@mail.gmail.com> <200911121529.nACFT8ad037531@lava.sentex.ca>
next in thread | previous in thread | raw e-mail | index | archive | help
The storm threshold is a tuneable, like for 10G the default (1000) is guaranteed to be too low, we usually up it to 10000 in our lab. In this case its not indicative of a problem so just increase until it does not throttle the source. Jack On Thu, Nov 12, 2009 at 7:28 AM, Mike Tancsa <mike@sentex.net> wrote: > At 05:19 PM 11/11/2009, Jack Vogel wrote: > > Ya, so maybe a more robust solution to that failure would be a good >> thing, but this is not the time for more elaborate code, having this small >> fix put in is low risk and the better failure response can come later. >> >> In fact, I have a number of places in the ixgbe driver where I'm pondering >> over what to do in failure mode anyway. >> > > [trimming re@freebsd as this isnt a release issue] > > No panics yet, but at around 350k pps, I see > > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > interrupt storm detected on "irq257:"; throttling interrupt source > > 0(ich10)# vmstat -i > interrupt total rate > irq16: uhci0+ 22 0 > irq18: ehci0 uhci5 15 0 > irq19: fwohci0++ 78 0 > irq21: uhci1 17 0 > irq23: uhci3 ehci1 2 0 > cpu0: timer 175404021 1998 > irq256: igb0 19731770 224 > irq257: igb0 30741918 350 > irq258: igb0 14 0 > irq259: igb1 71514077 814 > irq260: igb1 28098687 320 > irq261: igb1 2 0 > irq262: em0 4435336 50 > irq263: ahci0 89527 1 > cpu7: timer 175395421 1998 > cpu6: timer 175394148 1998 > cpu1: timer 175395011 1998 > cpu4: timer 175394583 1998 > cpu5: timer 175395129 1998 > cpu3: timer 175395466 1998 > cpu2: timer 175394884 1998 > Total 1557780128 17749 > 0(ich10)# > > > > > > -------------------------------------------------------------------- > Mike Tancsa, tel +1 519 651 3400 > Sentex Communications, mike@sentex.net > Providing Internet since 1994 www.sentex.net > Cambridge, Ontario Canada www.sentex.net/mike > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2a41acea0911120938g60de3ff0o6fbca1c06bdeb7a3>