xfs
[Top] [All Lists]

RE: Linux + XFS + SCSI = Problems?

To: "'Charles Radeke'" <charles.radeke@xxxxxxxxxxxxxxxxxx>, "Gonyou, Austin" <austin@xxxxxxxxxxxxxxx>, linux-xfs@xxxxxxxxxxx
Subject: RE: Linux + XFS + SCSI = Problems?
From: "Gonyou, Austin" <austin@xxxxxxxxxxxxxxx>
Date: Fri, 2 Nov 2001 11:35:54 -0600
Sender: owner-linux-xfs@xxxxxxxxxxx
Thanks for the reply Charles. Viele Danke! Anyway, I have concerns which are
aligned with what you 
are speaking of. It is a difficult thing to understand as well since I can
make the problem happen at will, 
but there is not traceable data on it. Strace won't show me anything. Is
there something else I could do, while
watching my process run away, that could help me diagnose this? Thanks
again.

-- 
Austin Gonyou
Systems Architect, CCNA
Coremetrics, Inc.
Phone: 512-796-9023
email: austin@xxxxxxxxxxxxxxx 

> -----Original Message-----
> From: Charles Radeke [mailto:charles.radeke@xxxxxxxxxxxxxxxxxx]
> Sent: Friday, November 02, 2001 4:31 AM
> To: Gonyou, Austin; linux-xfs@xxxxxxxxxxx
> Subject: Re: Linux + XFS + SCSI = Problems?
> 
> 
> "Gonyou, Austin" wrote:
> 
> > Thanks VERY much for the response. I sure do appreciate it. 
> That is the one
> > thing I've not done yet, (removed smp), but I hope to do it 
> soon. Of note,
> > it seems to be XFS + SMP, but not EXT2 + SMP, or REISER+ 
> SMP. That much I
> > know so far.
> > Austin
> >
> 
> yes, I used reiserfs over a year before I changed (of course 
> 8-)) to xfs..
> also at SMP machine and I never had problems as described... 
> possibly xfs is
> not the basic reason.. I read messegas about 99% cpu running 
> kupdated at a
> time where xfs was not released... I think more that xfs 
> points out a problem
> deep in the kernel source.. someone of the freaks will solve 
> it next time I
> hope, because no one wil run xfs on a SMP server in a view to 
> such problem..
> xfs filesystem was mountable after that but files created, 
> changed or opened
> after the kupdated starts the long run are completely 
> desapeared or filled
> with "@" after reboot.. very strange situation. that means 
> all you or the
> system do after the bug occure is complete sensless, no 
> aktion is really done
> with the filesystem but you can maybe untar 400MB, go in, and compile
> successfull, make install, or anyrthing, you can work with 
> the installed app;
> BUT if you reboot all ist lost.
> 
> c. radeke
> 


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