LinHES Forums
http://forums.linhes.org/

Did I mess everything up?
http://forums.linhes.org/viewtopic.php?f=11&t=14581
Page 1 of 1

Author:  dir3wolf [ Thu Mar 15, 2007 5:08 am ]
Post subject:  Did I mess everything up?

Hi,

I have been using knoppmyth for about a year now. I had some minor problems and every time I was able to fix them and continue to be happy user. I reinstalled every time new version was out, and everything was perfect.
But yesterday picture on one of my frontends froze so I went to my master backend to try end fix the problem. While trying to restart backend it froze too. Completely. I did a hard reset and all hell broke loose. I couldn't log in, and I've noticed error message during boot that mysql is not able to write to /var/lib/mysql because partition is full. I immediately remembered coming across posts here about some logs running wild. My backend is installed in a triple-boot with ubuntu and winxp (I hardly ever use them, but they are there).
I booted to ubuntu, mounted mythtv root partition and started poking around /var/log folder.
I think I was paying with interest for my year without problems because at the same time my ISP's link went down, and I couldn't consult wiki or forums for some advice. I was also struck with overwhelming stupidity combined with being very tired (it was around 2AM) and impatient so I thought what the hell I'll just delete everything in my /var/log dir and see what happens. Doing that I was able to get only 70MB of free space so I guess problem was not in logs.
I am not a newbie, but I am not a pro either. I know my way around, but I never had any serious problems with my linux machines before (that's why I switched all my machines to linux) and I know very little about logs - I am a user, not really an admin.
My question is - how much damage have I done in my moment of imapatient stupidity? What to do next? Can I patch things up? and where to look for the problem?
Now the worst part - I didn't run mythbackup for about a month.
Mea culpa, mea culpa, mea maxima culpa.

Author:  bruce_s01 [ Thu Mar 15, 2007 7:39 am ]
Post subject: 

I don't think you should have any real problems, I think the logs should be recreated when necessary. Did you delete the directories inside /var/log as well?
What version of KnoppMyth are you using?
What is the size of your KM root partition?
Do you use Mythsteam?

Bruce S.

Author:  dir3wolf [ Thu Mar 15, 2007 9:11 am ]
Post subject: 

Thanks Bruce,
I deleted everything including dirs.
I am using R5E50
My root partition is 6GB and I have separate partitions for /mythtv and recordings.
I don't use Mythstream.

Author:  bruce_s01 [ Fri Mar 16, 2007 5:46 am ]
Post subject: 

Thanks for the info.
Quote:
I deleted everything including dirs.
There might be some issues with the apps that write to subdirs.
Quote:
I am using R5E50
Good, there was an issue with R5C7 not rotating log files.
Quote:
My root partition is 6GB and I have separate partitions for /mythtv and recordings.
Thats bigger than usual, so I can't see where you would run out of space.
Quote:
I don't use Mythstream.
In the earlier versions, the streamed files would be under the mythtv home directory, so would occupy space on the / partition, this has now been symlinked to a directory on the myth partition.

What I would suggest is having a look at /var/log and check if the subdirectories are recreated, if not you will need to recreate the following directories with the correct ownership and permissions. I would recommend halting the backend and apache to reduce any problems while the directories are absent.
Code:
 ls -l /var/log |grep dr
drwxr-xr-x 2 root            root              4096 Mar 11 06:26 apache2
drwxr-xr-x 2 root            root              4096 Oct 30 06:54 fsck
drwxr-xr-x 2 Debian-ipw3945d Debian-ipw3945d   4096 Nov 22 06:13 ipw3945d
drwxr-s--- 2 mysql           adm               4096 Mar 16 06:25 mysql
drwxrwsr-x 2 mythtv          mythtv            4096 Mar 16 06:25 mythtv
drwxr-sr-x 2 news            news              4096 Dec 29 11:47 news
drwxr-xr-x 2 ntp             ntp               4096 Mar 16 06:25 ntpstats
drwxr-x--- 3 root            adm               4096 Mar 15 06:25 samba
drwxr-xr-x 2 root            root              4096 Aug 27  2006 sysstat


Hope this helps.

Bruce S.

Author:  dir3wolf [ Sat Mar 17, 2007 6:28 am ]
Post subject: 

bruce_s01 wrote:
What I would suggest is having a look at /var/log and check if the subdirectories are recreated, if not you will need to recreate the following directories with the correct ownership and permissions. I would recommend halting the backend and apache to reduce any problems while the directories are absent.
Code:
 ls -l /var/log |grep dr
drwxr-xr-x 2 root            root              4096 Mar 11 06:26 apache2
drwxr-xr-x 2 root            root              4096 Oct 30 06:54 fsck
drwxr-xr-x 2 Debian-ipw3945d Debian-ipw3945d   4096 Nov 22 06:13 ipw3945d
drwxr-s--- 2 mysql           adm               4096 Mar 16 06:25 mysql
drwxrwsr-x 2 mythtv          mythtv            4096 Mar 16 06:25 mythtv
drwxr-sr-x 2 news            news              4096 Dec 29 11:47 news
drwxr-xr-x 2 ntp             ntp               4096 Mar 16 06:25 ntpstats
drwxr-x--- 3 root            adm               4096 Mar 15 06:25 samba
drwxr-xr-x 2 root            root              4096 Aug 27  2006 sysstat


Hope this helps.

Bruce S.


Everything seems to work now. Thanks Bruce.
The only problem now is that mythbackup shows this at the end:

Code:
Warning, table lists are not identical!

The backup is bad or already out of date!


but I don't know is that related

Author:  tjc [ Sat Mar 17, 2007 10:43 am ]
Post subject: 

dir3wolf wrote:
The only problem now is that mythbackup shows this at the end:

Code:
Warning, table lists are not identical!

The backup is bad or already out of date!


but I don't know is that related

What differences does it show right above that? This usually indicates that something is still diddling with the DB despite the backups efforts to stop all the things which could interfer. You're not using the frontend at the same time or running one of the babysitting scripts that restart the backend are you? There are also a couple "harmless" differences that can occur here, but it's hard to say without the messages.

Author:  dir3wolf [ Sat Mar 17, 2007 4:49 pm ]
Post subject: 

tjc wrote:
What differences does it show right above that? This usually indicates that something is still diddling with the DB despite the backups efforts to stop all the things which could interfer. You're not using the frontend at the same time or running one of the babysitting scripts that restart the backend are you? There are also a couple "harmless" differences that can occur here, but it's hard to say without the messages.


Next time I tried to run mythbackup it couldn't complete because /myth ran out of free space. This time I was able to identify perpertrator. It was .xsession-errors which grew to 3GB. I deleted it and mythbackup finished without errors.
How can I prevent .xsession-errors from growing to big?

Page 1 of 1 All times are UTC - 6 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/