[OpenAFS] Re: [OpenAFS-devel] Re: OpenAFS bug, Cache gets full and is not flushed (fwd)

[Derek Atkins]

You can try back-porting the 2.4.8 support into OpenAFS 1.0.4 and try
it..  In particular, make sure that struct vnode in
src/afs/LINUX/osi_vfs.h matches struct inode in
linux-2.4.8/include/linux/fs.h

-derek

Patrick Sodre Carlos <sodre@wam.umd.edu> writes:

> This does not make sense since I was able to copy the same file
> before(openafs 1.1.4). But Now that I've upgraded to kernel 2.4.8 I
> "have" to use openafs 1.2.0.
> 
> Hoping for other ideas,
> Patrick Sodre Carlos
> 
>  On Fri, 14 Sep 2001, Shyh-Wei Luan wrote:
> 
> > 
> > I believe you need to have your client machine's cache partition size as
> > large as the largest file that you are "writing" into the AFS space through
> > this machine.   AFS client does "write on close" and all the file data
> > stays in the client cache until the file is closed or when fsync is called.
> > 
> > Shyh-Wei Luan
> > 
> > Patrick Sodre Carlos <sodre@wam.umd.edu>@openafs.org on 09/14/2001 10:10:18
> > AM
> > 
> > Sent by:    openafs-devel-admin@openafs.org
> > 
> > 
> > To:    openafs-info@openafs.org, openafs-devel@openafs.org
> > cc:
> > Subject:    [OpenAFS-devel] Re: OpenAFS bug, Cache gets full and is not
> >        flushed (fwd)
> > 
> > 
> > 
> > 
> > Hey Guys,
> > 
> > The file I was trying to copy was the size of a CD, 640MB I suppose.
> > 
> > Patrick Sodre
> > ---------- Forwarded message ----------
> > Date: 14 Sep 2001 09:06:17 -0400
> > From: Derek Atkins <warlord@MIT.EDU>
> > To: Patrick Sodre Carlos <sodre@wam.umd.edu>
> > Subject: Re: OpenAFS bug, Cache gets full and is not flushed
> > 
> > Oh, you should DEFINITELY be sending this to openafs-info or
> > openafs-devel (@openafs.org).  I don't know why you decided to send
> > this directly to me.
> > 
> > How big is this file that you're trying to copy?
> > 
> > -derek
> > 
> > Patrick Sodre Carlos <sodre@wam.umd.edu> writes:
> > 
> > > Due to the above, the cell stops working and any progress after to finish
> > > copying the files after that is "halted".
> > >
> > > Data of my system:
> > >  - Kernel 2.4.8 (SMP 2CPUS)
> > >  - OpenAFS 1.2.0
> > >
> > > How did this bug happens:
> > >  - Set the cache to live in a separate partition of 200 Megabytes
> > >  - Get a big file to copy to AFS (Mandrake.iso)
> > >  - Copy the Mandrake.iso file to the /afs/.cellname/somevolume
> > >
> > > Expected:
> > >  - For the whole file to be copied since there was "plenty" of space
> > > available in the volume.
> > >
> > > Result:
> > >  - The "cp"/"rsync"/"ftp" halts as soon as the cache is filled
> > > (180Mega) and I can't get control of the system or "kill -9", thefore a
> > > reboot it necessary.
> > >
> > >
> > > Thanks in advance,
> > >
> > > Patrick Sodre
> > > PS: Tell me if I should be sending this message somewhereelse
> > >
> > >
> > 
> > --
> >        Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
> >        Member, MIT Student Information Processing Board  (SIPB)
> >        URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
> >        warlord@MIT.EDU                        PGP key available
> > 
> > _______________________________________________
> > OpenAFS-devel mailing list
> > OpenAFS-devel@openafs.org
> > https://lists.openafs.org/mailman/listinfo/openafs-devel
> > 
> > 
> 
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info

-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord@MIT.EDU                        PGP key available