Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 23 May 2010 21:54:36 -0500
From:      Brandon Gooch <jamesbrandongooch@gmail.com>
To:        Kurt Jaeger <pi@opsec.eu>
Cc:        freebsd-net@freebsd.org, Kip Macy <kip.macy@gmail.com>
Subject:   Re: kern/146792: [flowtable] flowcleaner 100% cpu's core load
Message-ID:  <AANLkTikXrtkLUswmjRpmAvWOQqOj-snwp4cYbBUUgR0d@mail.gmail.com>
In-Reply-To: <201005232206.o4NM6gor016873@freefall.freebsd.org>
References:  <201005232206.o4NM6gor016873@freefall.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, May 23, 2010 at 5:06 PM, Kurt Jaeger <pi@opsec.eu> wrote:
> The following reply was made to PR kern/146792; it has been noted by GNAT=
S.
>
> From: Kurt Jaeger <pi@opsec.eu>
> To: bug-followup@FreeBSD.org, niko@gtelecom.ru
> Cc:
> Subject: Re: kern/146792: [flowtable] flowcleaner 100% cpu's core load
> Date: Sun, 23 May 2010 16:19:25 +0200
>
> =A0Hi!
>
> =A0I observe a similar behaviour on a 8.0-RELEASE-p2 i386 GENERIC
> =A0kernel.
>
> =A0System receives 2 BGP4 fullfeeds (approx. 310K routes each).
>
> =A0The system is still running, a few processes are unkillable or
> =A0die only after a long amount (1-2h) of time.
>
> =A0Here's the list of unkillable processes:
>
> =A080871 =A0?? =A0R =A0 =A0 =A00:00.00 /bin/sh /etc/periodic/daily/470.st=
atus-named
> =A076499 =A0?? =A0Rs =A0 =A0 0:00.01 sshd: [accepted] (sshd)
> =A076922 =A0?? =A0Rs =A0 =A0 0:00.01 sshd: [accepted] (sshd)
>
> =A0flowcleaner looks pretty busy (for an uptime of approx. 40h):
>
> =A0 =A022 =A0?? =A0RL =A0 1209:50.98 [flowcleaner]
>
> =A04:17PM =A0up 1 day, 22:22, 2 users, load averages: 7.20, 6.53, 5.81
>
> =A0quagga is running on the system, bgpd mgmt cli is no longer reachable:
>
> =A0# telnet 0 2605
> =A0Trying 0.0.0.0...
> =A0Connected to 0.
> =A0Escape character is '^]'.
>
> =A0^]
> =A0telnet> close
> =A0Connection closed.
> =A0#
>
> =A0What can I do to help to debug this ?
> =A0No console access available right now, but can probably made available=
.
>
> =A0This is a production host, but not yet super-critical, so...

I know absolutely nothing about quagga, and very, very little about
the flowcleaner process (or flowtable, no man page), but I DO KNOW
that Kip Macy suggested disabling:

options FLOWTABLE

from the kernel config of the machine experiencing the issue. This was
back in December 2009, so I'm not sure about a resolution to the
actual problem (or if it is just inherent in the design of the per-cpu
routing cache).

Perhaps Kip may have more insight?

-Brandon



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTikXrtkLUswmjRpmAvWOQqOj-snwp4cYbBUUgR0d>