Re: bdflush/rpciod high CPU utilization, profile does not make sense

[Jakob Oestergaard <jakob@unthought.net>]

On Mon, Apr 11, 2005 at 10:35:25AM -0400, Trond Myklebust wrote:
> må den 11.04.2005 Klokka 15:47 (+0200) skreiv Jakob Oestergaard:
> 
> > Certainly;
> > 
> > http://unthought.net/binary.dmp.bz2
> > 
> > I got an 'invalid snaplen' with the 90000 you suggested, the above dump
> > is done with 9000 - if you need another snaplen please just let me know.
> 
> So, the RPC itself looks good, but it also looks as if after a while you
> are running into some heavy retransmission problems with TCP too (at the
> TCP level now, instead of at the RPC level). When you get into that
> mode, it looks as if every 2nd or 3rd TCP segment being sent from the
> client is being lost...

Odd...

I'm really sorry for using your time if this ends up being just a
networking problem.

> That can mean either that the server is dropping fragments, or that the
> client is dropping the replies. Can you generate a similar tcpdump on
> the server?

Certainly;  http://unthought.net/sparrow.dmp.bz2


-- 

 / jakob