[rhelv6-beta-list] network errors during kickstart

[Simon Matter]

> Hello Nico,
>
> The same webserver is serving a lot of other stuff with no problems. Also
> when installing Fedora13 off the same webserver and onto the same machine,
> no problems. Parallel installs of RHEL6b2 and F13, the former does have
> the
> problem, the latter not.

Just a guess, there were some changes in curl which is also used by yum.
See here https://bugzilla.redhat.com/show_bug.cgi?id=589132
Maybe it's related somehow?

Simon

>
> The filesystem holding the installation data is a GFS filesystem, serving
> lots of other stuff without problems. If the files could not be read, I'd
> expect errors in the webserver log, but there are none. The files accessed
> actually are all symlinks to real files on the same filesystem (same
> problem
> when accessing the physical files, rather than the symlinks), but again,
> doing the exactly same thing for F13 and older versions of RHEL6 cause no
> problems, RHEL6b2 does have the problem, for both physical and virtual
> installs.
>
> Cheers,
>
> Kurt
>
> On 7 July 2010 12:40, Nico Kadel-Garcia <nkadel at gmail.com> wrote:
>
>> On Wed, Jul 7, 2010 at 4:10 AM, Kurt Keller <kkathag at gmail.com> wrote:
>> > Since beta 2 we encounter one error on every kickstart install. This
>> > prevents us from performing the fully automated installs we require,
>> as
>> > somebody has to manually select 'retry'.
>> > Has anybody else these errors?
>> > Is there any switch which could be used to automatically retry instead
>> of
>> > waiting for an admin to select 'retry'?
>> >
>> > What happens is:
>> > - we PXE boot
>> > - kickstart and channels are accessed via http
>> > - while in the stage "Starting installation process", there will
>> always
>> be
>> > exactly one rpm package (not always the same) with the following
>> error:
>> "The
>> > file xxxx.rpm cannot be opened. This is due to a missing file, a
>> corrupt
>> > package or corrupt media..."
>> > - selecting retry works fine and the installation continues
>> > - 30 seconds before this error occurs, we see in the webserver logs,
>> that
>> > last request for partial of this file (status code 206)
>> > - at the time this occurrs we see syslog messages from the machine
>> being
>> > installed saying it timed out on the file
>> >
>> > We don't think this is a problem with the webserver, network or the
>> like,
>> as
>> > it happens always with RHEL6b2 and never with RHEL4/5 F13, using the
>> same
>> > installation infrastructure and even running at the same time.
>> >
>> > Thanks,
>> >
>> > Kurt
>>
>> Interesting. Is your webserver funky in some way? Overwhelmed, for
>> example? Or having difficulties with a shared filesystem that is
>> storing the repository?
>>
>> _______________________________________________
>> rhelv6-beta-list mailing list
>> rhelv6-beta-list at redhat.com
>> https://www.redhat.com/mailman/listinfo/rhelv6-beta-list
>>
> _______________________________________________
> rhelv6-beta-list mailing list
> rhelv6-beta-list at redhat.com
> https://www.redhat.com/mailman/listinfo/rhelv6-beta-list