Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 01 Sep 2006 20:59:53 -0700
From:      Sam Leffler <sam@errno.com>
To:        dandee@volny.cz
Cc:        freebsd-stable@freebsd.org, current@freebsd.org
Subject:   Re: atheros driver under high load, panics and even more freezes
Message-ID:  <44F901B9.3040309@errno.com>
In-Reply-To: <001e01c6ce2e$1253ff30$6508280a@tocnet28.jspoj.czf>
References:  <001e01c6ce2e$1253ff30$6508280a@tocnet28.jspoj.czf>

next in thread | previous in thread | raw e-mail | index | archive | help
Daniel Dvoøák wrote:
> Hi all,
>  
> first of all, I´m sorry maybe for my bad English.
>  
> We have 2 routers which I maintain in our mesh wireless community network.
>  
> The Router 1 has 2 atheros adapters, ath0=wistron cm9, ath1=wistron cm10, of
> course some sisX, fxpX and so on.
> The Router 2 has 1 atheros adapter, ath1=wistron CM10.
>  
> My R1 panics and even more it freezes very often. Maybe the reason for
> panicing and freezing is the same and maybe not.
>  
> I started  (only after vmcore.5, so vmcore.6 is with this option)  to use
> "option SW_WATCHDOG" in both my custom kernels on the R1 and R2 recently in
> hope, it is some walkaround for freezing at least if not for panicing. 
>  
> This router was installed on the 1st of April 2006.
>  
> Statistics:
>  
> 9 panics with 8 kernel dumps, 1 missed
>  
> 10 freezes
>  
> I think that all panics some how connected to athX taskq process, page fault
> in kernel panic and sbflush_locked.

Why?

>  
> I guess that panic comes when router transmits and receives datas at the
> maximum throughput for setted nominal media rate speed, exactly 24Mbps, more
> I do not use, because there are problems with quagga 
>  
> ospfd packets, it is known issue.
>  
> Today I did a small test with throughput.
>  
> Router 1 executed this command:
>  
> # ping -i 0.001 -c 100000 -s 1472 ANY IP
>  
> As you see, it is not even flood ping, it is almost flood, but not flood.
>  
> Throughput was about 1,13-1,2 MB/s as bmon showed me. I notice there is not
> any qos and icmp.limit is so high net.inet.icmp.icmplim: 2147483647
> net.ineticmp.icmplim_output: 0.
> 
>  
> First 5 s latency was about 1,1-1,7 ms
> After it goes to 10-30, 50-70, 110-130, 270-300, up 300ms and packet loss
>  
> .... some seconds ....
>  
> panic

	<...lots of stuff deleted...>

Sounds like a resource leak to me.  You've got crash dumps, look at
memory usage with vmstat and/or netstat.  Past that it sounds like
you're running 6.1 RELEASE which is now 6+ months old.  Many bugs have
been fixed including, I believe, some resource-related ones.  Please try
6-STABLE.

	Sam




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?44F901B9.3040309>