On 6/30/12 10:05 PM, Bob Cochran wrote:
On 6/29/12 11:19 PM, James Cammarata wrote:
> On Fri, Jun 29, 2012 at 10:07 PM, Parthasarathy, Balaji (BCS,
> Cupertino, USA)<balaji.parthasarathy(a)hp.com> wrote:
>>> -----Original Message-----
>>> From: cobbler-bounces(a)lists.fedorahosted.org [mailto:cobbler-
>>> bounces(a)lists.fedorahosted.org] On Behalf Of Bob Cochran
>>> Sent: Friday, June 29, 2012 6:14 PM
>>> To: cobbler-list
>>> Subject: [cobbler] Another question about 'cobbler reposync'
>>>
>>> Thanks to much help from you great people, my 'cobbler reposync'
>>> started
>>> progressing well...until now, it seems stalled at this point:
>>>
>>> running: createrepo -g
>>> repodata/42ec1be1745e71753d57892fefff57fc16d98541662637e1a8f2c0b2f188
>>> 70f6-comps-f17.xml
>>> --deltas -c cache -s sha
>>> /var/www/cobbler/repo_mirror/f17-x86_64-everything
>>> 13/27033 - Packages/q/qalculreamer-devel-0.10.2-1.fc17.i686.rpm
>>>
>>> Does it just take a long time for the above step to complete or is
>>> it stuck on
>>> qalculreamer-devel? I would love to see it progress from package
>>> 13/27033 to
>>> package 27033/27033!
>> Yes, createrepo does take some time. I have seen it take about 1-2
>> mins for indexing ~2000 RPMs. Since there are 27000+ RPM packages in
>> your case, it'll take longer.
>>
> You can also do something like "sudo lsof | grep createrepo" to see
> what file it's actually working on at the moment.
Thank you both! I was (yesterday) getting something like 'unhandled
exception in cobblerd' messages in /var/log/messages while running
'createrepo', but my efforts to do a successful 'cobbler reposync'
action have been hampered by thunderstorms which caused massive power
outages in my area. (I live close to Washington, D. C.) The reposync
process seemed to stall and stop dead after the 'unhandled exception'
messages. If this condition persists, I will post the exact text of
/var/log/messages.
Now that power is back on again, to restart the process, I deleted the
/var/www/cobbler/repo_mirror/f17-x86_64-everything directory and I am
now rerunning 'cobbler reposync'. Assuming power stays on tonight, we
will see what happens in the morning.
If the above is not the correct way to restart an apparently stalled
reposync process, please let me know the procedure. I'm quite new at
this.
Thanks again so much!
Bob Cochran
[The next morning] it looks like the 'createrepo' step keeps failing.
With more investigation, I am fairly sure this is because the hard drive
I am using is failing. I am now getting floods of kernel messages about
bad blocks that didn't show up before. So it looks to me like the best
thing to do is to clone this existing hard drive to a new hard drive
before it fails totally. Replace the failed hard drive with the clone
drive. Then rerun 'cobbler reposync'.
Bob