On Thu, Mar 09, 2006 at 03:11:42PM -0800, Bernard Li wrote:
> Hi Ben:
> 
> I just re-read your email, you did mention that it hangs even if the
> filesize is 0 bytes.
> 
> Anyways, the only time I have encountered similar issue is when there
> are some network/firewall problems (eg. pfilter on OSCAR) - perhaps
> you can check that out?

These machines are on a virgin network.  The image server has port 22
open and little else.  The clients are configured to tunnel rysnc
through ssh (using the --ssh-user flag to si_updateclient).  There are
no particular firewalls or filters in between the two machines - they're
all connected directly through a gigabit switch.

Honestly, I havn't spent too much time fighting this one since it only
seems to happen when completely re-indexing a machine.  Incremental
updates proceed normally.  But it is frustrating...

Does my guess that it may have something to do with the running
environment getting wiped out from under the updateclient script have
any merit?  The complete reimage wipes out the kernel, all the binaries,
including the updateclient script itself...  but of course, it also
fills everything back in before the rsync finishes...

-ben

> 
> Cheers,
> 
> Bernard
> 
> 
> -----Original Message-----
> From: [EMAIL PROTECTED] on behalf of Ben Hartshorne
> Sent: Thu 09/03/2006 14:16
> To: [email protected]
> Subject: Re: [Sisuite-users] Anyone know why si_getimage would fail on Fedora 
> Core 4 x64?
>  
> Bernard,
> 
> I wasn't totally clear in my initial message - the problem occurs for me
> even after zero-ing the lastlog file.  And if I straight-out delete the
> lastlog file, it hangs on the next file, whatever that may be.
> 
> -ben
> 
> 
> On Thu, Mar 09, 2006 at 12:40:35PM -0800, Bernard Li wrote:
> > Hi Drew:
> > 
> > I think Ben hit it with a nail there...  on a Red Hat 64-bit machine
> > (not confirmed on other 64-bit OSes...) the /var/log/lastlog is
> > huge...  this apparently is a known bug with the OS and I am not sure
> > if it is fixed yet.  Anyways, in OSCAR, we fixed it by zeroing the
> > file - take a look at the following bug we filed for OSCAR:
> > 
> > http://sourceforge.net/tracker/?group_id=9368&atid=109368&func=detail&aid=1266306
> > 
> > Cheers,
> > 
> > Bernard
> > 
> > 
> > -----Original Message-----
> > From: [EMAIL PROTECTED] on behalf of Drew Weaver
> > Sent: Thu 09/03/2006 11:25
> > To: [email protected]
> > Subject: RE: [Sisuite-users] Anyone know why si_getimage would fail on 
> > Fedora Core 4 x64?
> >  
> > It only does it on FC4_64 to me, every other OS completes fine, and I
> > mean every other OS lol
> > 
> > -----Original Message-----
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED] On Behalf Of Ben
> > Hartshorne
> > Sent: Thursday, March 09, 2006 1:53 PM
> > To: [email protected]
> > Subject: Re: [Sisuite-users] Anyone know why si_getimage would fail on
> > Fedora Core 4 x64?
> > 
> > On Thu, Mar 09, 2006 at 12:39:40PM -0500, Drew Weaver wrote:
> > > There aren't any, it literally hangs after it copies the last file in
> > > the /var/log directory forever.
> > 
> > Drew,
> > 
> > I think I see something similar, but mine hangs before getting to the
> > last file.  It always hangs at the same place.  If I stop it and then
> > run it again, it goes to completion.
> > 
> > I see the failure when doing a massive re-image (for example, dumping a
> > FC4 over FC3 or dumping a CentOS image on a Fedora server, etc.) and
> > assumed that it had something to do with deleting the files that the
> > program is using to run, but I havn't been able to track it down either.
> > 
> > More details on the symptoms:
> > Before the hang, disk activity and processor are both high (I'm going
> > through ssh).  After the hang, disk and CPU both drop to 100% idle.  It
> > doesn't time out (up to several hours).  processes appear to be fine
> > (not zombied or in some other way spinning).  If you delete the file at
> > which it hangs, it hangs at the next one.  the file mine was hanging on
> > was /var/log/lastlog (hangs there even when the file is 0 bytes) or
> > /var/log/dmesg (after I deleted a bunch of log files).  There isn't
> > anything particularyl interesting (large size or whatnot) about either
> > the file that came before it or the file I guess will come after it.  
> > 
> > For me, at least, the problem only seems to occur when I completely
> > re-image the machine, and does not occur on incremental updates.
> > 
> > ::sigh::
> > 
> > -ben
> > 
> > > 
> > >  
> > > 
> > > -Drew
> > > 
> > >  
> > > 
> > > ________________________________
> > > 
> > > From: [EMAIL PROTECTED]
> > > [mailto:[EMAIL PROTECTED] On Behalf Of
> > Bernard
> > > Li
> > > Sent: Thursday, March 09, 2006 12:28 PM
> > > To: [email protected];
> > > [email protected]
> > > Subject: RE: [Sisuite-users] Anyone know why si_getimage would fail on
> > > Fedora Core 4 x64?
> > > 
> > >  
> > > 
> > > It would help if you provide more error messages for us to
> > troubleshoot
> > > the issue.
> > > 
> > > Cheers,
> > > 
> > > Bernard
> > > 
> > > 
> > > -----Original Message-----
> > > From: [EMAIL PROTECTED] on behalf of Drew
> > Weaver
> > > Sent: Wed 08/03/2006 18:41
> > > To: [email protected]
> > > Subject: [Sisuite-users] Anyone know why si_getimage would fail on
> > > Fedora Core 4 x64?
> > > 
> > >             It dies after rsync copies the last file to the server
> > from
> > > the client. The client is fully updated im running 3.6.2
> > > 
> > > 
> > > 
> > > -Drew
> > > 
> > > 
> > > 
> > 
> > -- 
> > Ben Hartshorne
> > email: [EMAIL PROTECTED]
> > http://ben.hartshorne.net
> > 
> > 
> > -------------------------------------------------------
> > This SF.Net email is sponsored by xPML, a groundbreaking scripting language
> > that extends applications into web and mobile media. Attend the live webcast
> > and join the prime developer group breaking into this new coding territory!
> > http://sel.as-us.falkag.net/sel?cmd=k&kid0944&bid$1720&dat1642
> > _______________________________________________
> > Sisuite-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/sisuite-users
> > 
> 
> -- 
> Ben Hartshorne
> email: [EMAIL PROTECTED]
> http://ben.hartshorne.net
> 

-- 
Ben Hartshorne
email: [EMAIL PROTECTED]
http://ben.hartshorne.net

Attachment: signature.asc
Description: Digital signature

Reply via email to