xfs
[Top] [All Lists]

Re: TAKE - Delay Buffer Changes

To: Ananth Ananthanarayanan <ananth@xxxxxxxxxxxxxxxxx>
Subject: Re: TAKE - Delay Buffer Changes
From: Danny <danscox@xxxxxxxxxxxxxx>
Date: Wed, 28 Feb 2001 08:13:38 -0500
Cc: Linux-XFS <linux-xfs@xxxxxxxxxxx>
Organization: Connex Inc
References: <200102270040.f1R0eAV02973@xxxxxxxxxxxxxxxxx>
Reply-to: dcox@xxxxxxxxxx
Sender: owner-linux-xfs@xxxxxxxxxxx
Ananth,

Ananth Ananthanarayanan wrote:
> 
> Significant part of this change is to employ standard
> linux IO daemons & paths to handle delalloc pages by
> attaching buffers to all pages.  This also enables
> handling of sync()-type operation (changes yet to come) ...

        I just wanted everyone to know: this change has enabled me to run two
Bonnies overnight with XFS on RAID1.  The machine DID become unusable at
one point yesterday.  I had a shell prompt, but any command other than a
shell built-in immediatly got a core dump(!).  After that, however, it
ran just fine.  From about 1:30 PM through now (~ 8:00 AM), the two
Bonnies have been running great.  I've been having some RAM problems
too, and this "hang" may have been due to that.  I've NOT seen the
"0-order allocation failed" messages, which was my main sticking point.

        Also, either because I'm using the "-o logbufs=4,logbsize=32768" mount
option, or the recent changes, or both, the Bonnie block writes went
from 14+ MB/sec to 20+, and the block reads went from 9+ MB/sec to 10+.

        Yesterday I was "cautiously optimistic".  Today, I'm "comfortably
optimistic".

        Great work, Ananth!  Thanks!

-- 
"Men occasionally stumble over the truth, but most of them pick 
themselves up and hurry off as if nothing had happened." 
   -- Winston Churchill 

Danny

<Prev in Thread] Current Thread [Next in Thread>