Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 12 May 2003 13:39:22 -0700 (PDT)
From:      Don Lewis <truckman@FreeBSD.org>
To:        rwatson@FreeBSD.org
Cc:        current@FreeBSD.org
Subject:   Re: rpc.lockd spinning; much breakage
Message-ID:  <200305122039.h4CKdMM7048544@gw.catspoiler.org>
In-Reply-To: <Pine.NEB.3.96L.1030512100945.76804F-100000@fledge.watson.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 12 May, Robert Watson wrote:

> On the server side, rpc.lockd is unusually busy as a result:
> 
>   PID USERNAME   PRI NICE   SIZE    RES STATE    TIME   WCPU    CPU COMMAND
> 47547 root       121    0  1592K  1224K RUN    448:30 98.68% 98.68% rpc.lockd
> 
> Usually the behavior that upsets the client is to run:
> 
> while (1)
>     periodic daily
>     periodic weekly
>     periodic monthly
> end
> 
> The rpc.lockd process remains extremely busy even after crash2 is rebooted
> and the stream of packets is no longer present.
> 
> I'm not sure how to go about debugging these problems, but the current
> scenario basically means I can't get both the crash boxes through their
> daily events if both the client and server are very busy (i.e., if they
> both run their daily events at the same time).  I'm going to reboot cboss
> and the systems and see if that flushes whatever nasty state hangs around,
> but any advice on the debugging process would be helpful.  Is there a way
> to get rpc.lockd on the server to dump it's state to a file?

Why not attach the process in gdb and step through the code to find the
loop?



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200305122039.h4CKdMM7048544>