Dear All,
I've got a small server at work, built from a Via Epia SP 8000E
mini-ITX motherboard and a Seagate SATA hard drive. The application
of this server is to provide versioned file storage, using subversion.
I'm running CentOS, with Apache 2 and SSL to provide controlled
access to the subversion repositories.
There have been two problems with this machine:-
Problem 1
Every once in a while one of the repositories gets corrupted, which
produces an error when attempting to use it with tortoisesvn or svn
from another Linux box. I then have to go in as root on ssh and do
svnadmin recover followed by chown -R apache:apache just to be sure,
then it works again.
Problem 2
The server had been running continually for maybe 70 days without
problems apart from problem 1. A couple of weeks ago there was a
kernel update. I rebooted the machine to run the new kernel in case
it included security fixes, and at 23:00 on Thursday night it died.
When I came in on Friday morning it was sat there with the hard drive
permanently lit and not responding. I had to push the reset button to
get it going again. /var/log/messages entries from this incident follow:-
Feb 2 21:01:01 localhost crond(pam_unix)[1399]: session opened for
user root by (uid=0)
Feb 2 21:01:01 localhost crond(pam_unix)[1399]: session closed for user root
Feb 2 22:01:01 localhost crond(pam_unix)[1403]: session opened for
user root by (uid=0)
Feb 2 22:01:01 localhost crond(pam_unix)[1403]: session closed for user root
Feb 2 23:00:01 localhost crond(pam_unix)[1407]: session opened for
user root by (uid=0)
Feb 3 09:10:02 localhost syslogd 1.4.1: restart.
Feb 3 09:10:02 localhost syslog: syslogd startup succeeded
At 23:00 on Thursday crond would have been running a bash script to
do a daily dump of the repositories.
BTW the server is supplied by an APC UPS so there should be no issues
with mains power supply problems.
As I'm not running on Intel hardware, does anybody know of any issues
with running CentOS on Via motherboards with Eden processors?
Thanks.
Dave Fletcher
