xfs
[Top] [All Lists]

Re: Problems with many processes copying large directories acrossan XFS

To: Adrian Head <adrian.head@xxxxxxxxxxxxxxx>
Subject: Re: Problems with many processes copying large directories acrossan XFS volume.
From: Simon Matter <simon.matter@xxxxxxxxxxxxxxxx>
Date: Tue, 11 Sep 2001 08:11:23 +0200
>received: from mobile.sauter-bc.com (unknown [10.1.6.21]) by basel1.sauter-bc.com (Postfix) with ESMTP id D557857306; Tue, 11 Sep 2001 08:11:23 +0200 (CEST)
Cc: linux-xfs@xxxxxxxxxxx
Organization: Sauter AG, Basel
References: <D1F276F384A0D311A7C500A0C9E89B381980E4@xxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
Adrian Head schrieb:
> 
> Again - thanks Simon for your time.
> 
> I have done ten's of tests now and I am getting a strong feeling that it
> is the same problems described here from Steve Lord:
> http://marc.theaimsgroup.com/?l=linux-xfs&m=99980808004670&w=2
> 
> Or the start of the thread:
> http://marc.theaimsgroup.com/?l=linux-xfs&m=99919025308712&w=2
> 
> The reason I am almost convinced it is the same issue is that about
> every 3 attempts I have only bdflush running and everything else in
> deadlock. - I will have to follow the instructions and see what output I
> get from kdb.

So do I.

> 
> It appears that although they could only get it to crash with SCSI the
> problem affects both SCSI & IDE.
> 
> I was going to short-cut the process and just download the cvs tree -
> but for some reason the XFS site times-out at the moment.  I will try
> later.
> 
> When your machine hangs - have you noticed any trends in what processes
> are left running?

Unfortunately it disabled kdb and was not able to get a process list.
I'm still trying more tests and I will also try a current kernel if time
permits.

Simon

> 
> Adrian Head P/L
> 
> > -----Original Message-----
> > From: Simon Matter [SMTP:simon.matter@xxxxxxxxxxxxxxxx]
> > Sent: Tuesday, 11 September 2001 00:28
> > To:   Adrian Head
> > Cc:   linux-xfs@xxxxxxxxxxx
> > Subject:      Re: Problems with many processes copying large
> > directories acrossan XFS   volume.
> >
> >
> > I tried 20, 40 and 80 simultanous cp with no crash. Then I changed the
> > file tree and the new tree has ~280M small files with 100b-50kb size.
> > When using 60 cp jobs the machine died. I could ping it but nothing
> > more. No ssh, no console, no shutdown. I try some more tests tonight.
> > I
> > try the same with ext2 as well to make sure it's XFS and not Softraid.
> >
> > -Simon
> >
> >



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