I came home today to a locked up frontend. After a reboot, the frontend wouldn't come back up and I saw that /dev/hda1 filled itself up. I deleted some logs to clear a little space, rebooted, but still no dice...I received a bunch of unable to connect to database messages, specifically

Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock'

Figuring the db was dead, I did a mythrestore to try to fix the issue. That ended with a message "Warning, table lists are not identical! The restore failed or was already modified!"

Does anyone know what I can try next to get this thing back up and running?

_________________
Currently running: R5.5, HD5000 x 2, PVR150, Athlon 64 3000+, Chaintech VNF4, 1GB RAM, 2 x 250GB in LVM, MSI NX6200TC -> AA 9A60 -> HDTV

- Fix the "logrotate" problem. I just posted a link to the old posting on this earleir in the weekend so searching for that keyword should find it within minutes.

- Repair your database. See Cecil's sticky posting under hints and tips.

- Make sure that nothing is accessing the DB before doing the restore. If the the front end is running, and you have logging to the DB enabled, and the only difference shown is the "mythlog" table having a couple extra entries you can ignore it.

Thanks, I'll do some searches and try to repair.

If mythrestore failed, then is my db now completely corrupted or in the same state as it was before I ran mythrestore?

_________________
Currently running: R5.5, HD5000 x 2, PVR150, Athlon 64 3000+, Chaintech VNF4, 1GB RAM, 2 x 250GB in LVM, MSI NX6200TC -> AA 9A60 -> HDTV

Unfortunately, Cecil's repair post didn't help. myisamchk didn't even find any error, which is strange considering I can't start /etc/init.d/mysql - it just fails.

_________________
Currently running: R5.5, HD5000 x 2, PVR150, Athlon 64 3000+, Chaintech VNF4, 1GB RAM, 2 x 250GB in LVM, MSI NX6200TC -> AA 9A60 -> HDTV

I'd DL a new ISO, check the md5sum, burn a new CD on different media at low speed, boot with the testcd option, and retry the upgrade. Sounds like something funky happened on the way to the installer.

I did a KM upgrade of R5E50 again to wipe the slate clean. I did a mythrestore afterwards and although it said the restored had failed, I ended up with a usable, albeit old, version of my database so I didn't have to start from scratch.

_________________
Currently running: R5.5, HD5000 x 2, PVR150, Athlon 64 3000+, Chaintech VNF4, 1GB RAM, 2 x 250GB in LVM, MSI NX6200TC -> AA 9A60 -> HDTV

An upgrade already does a restore as part of the process. All you need to do is have a valid set of backup files available in /myth/backup before you enter the password to complete the phase2 setup.

Understood, but for some reason I had to do it manually that time around. There must have been an error or something that I overlooked.

_________________
Currently running: R5.5, HD5000 x 2, PVR150, Athlon 64 3000+, Chaintech VNF4, 1GB RAM, 2 x 250GB in LVM, MSI NX6200TC -> AA 9A60 -> HDTV

Did you happen to note what the errors were? There are a couple common ones which are more or less harmless. The checks done after a restore are mostly looking to see that all the files which it extracts from the backup are there., and if the record counts in the DB tables match. Sometimes when the mythbackend server is restarted it cleans out some old stuff...