[Top] [All Lists]

Re: Linux 2.4.22 XFS 1.3.1 reservation ran out

To: Eric Sandeen <sandeen@xxxxxxx>
Subject: Re: Linux 2.4.22 XFS 1.3.1 reservation ran out
From: Soumen Chakrabarti <soumen@xxxxxxxxxxxxxx>
Date: Tue, 02 Dec 2003 11:36:30 +0530
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <1070295327.28534.2.camel@stout.americas.sgi.com>
References: <Pine.LNX.4.44.0310270844480.9654-100000@stout.americas.sgi.com> <3FC6E5C4.2080009@cse.iitb.ac.in> <1070295327.28534.2.camel@stout.americas.sgi.com>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031014 Thunderbird/0.3
On Fri, 2003-11-28 at 00:05, Soumen Chakrabarti wrote:
I downloaded the patch around 10/15, from this URL:
which seems to indicate that even on that date, there was a patch available for 2.4.22 (maybe I looked where I should not).

The above patch version/snapshot has the problem I posted earlier. But with ikeep I have not hit a problem up to 300*1024 files. Could not experiment with my original RAID array because that's in production; I tested with a smaller loop dev.

BTW, I rolled back to kernel 2.4.21 and applied these two patches:
and everything is fine, I have tested with Bonnie++ up to 600,000 files.

As you know, sudden power failure can lead to zeroed out blocks in XFS. Is there any tuning option to reduce the cache flush time or force journaling of data as well as metadata?

Thanks a lot for your help!

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