All, I just recently had to rebuild my environment after a hard drive crash on Master backend. I installed fresh R5.5 but the backend is crashing daily with a segfault. In the messages file:





The last thing seen in the mythbackend.log:



I see the following messages after restarting the backend:




I've seen other posts where database issues caused the backend to segfault. It is not clear how I should repair this. Any suggestions or pointers to actual posts with some examples on how to correct this would be great.

I think I found out how to repair it from a mysql forum.



I did this an I can now select from the table. Hopefully this resolves the segfaults.

Well it is still getting a segfault and not sure why. This time when I restarted it there was no messages regarding anything corrupt in the database.

Anyone else ever experience this or any pointers on how to track it down?

It's not totally clear from your posts - did you replace the hard drive after your crash, or did you re-install to the same hard drive?

I replaced the drive and did a fresh install. I lost all of my previous recordings and the mythconverg database was created as part of the new install.

Try running the optimize_db.sh script as root. It uses the strongest and most robust recovery techniques available. If that fails you may need to recreate the DB from scratch.

Thanks tjc. mythfilldatabase is core dumping as well. I am still not 100% sure it is the database, but not sure what else it could be.

To me, it sounds like hard drive corruption affecting the database files. You might want to run a file system check and, if you find corruption again, try changing the hard drive cable.

Well, I did a repair on every table. I did this before tjc mentioned the optimize script. It has not segfaulted since. I added a "ulimit -c unlimited" to the backend rc script so if it does core again at least I can take a ponder at the stack.

If the stack is completely useful, I may need to build a debug version of mythbackend, but I am keeping my fingers crossed that repairing all the tables corrected the problem.

Well, I spoke to soon it is still getting segfaults. I did run the optimize script but that still did not correct the problem. Also mythfilldatabase is segfaulting as well so what is next? I added a ulimit -c unlimited to the backend init.d script but I am still not getting core dumps.

If I were to rebuild everything is there a safe way to get the data out of mysql without re-introducing the corruption back in?

Upgrading to the same version is the standard way of handling broken installations. Follow the hitchhiker's guide to upgrading under hints&tips.

Essentially you will be backing up, upgrading and then restoring.

Be sure the database and file systems are in good working order before you start, the way you already have.

Wuh, I had a dumping system once and it's not fun at all. In my case it was bad hw. Replacing it was a true joy, but the damage done to the database and file system still haunts me. The suggested procedure took care of the remaining instability issues, but I did have to delete quite a few destroyed files. This was in 2007, when the ext3 file system was discovered to be "unreliable", thus not coping well with system crashes. Hope you are luckier!

MythArchive is a good option for anything you can't afford to lose.

Cheers,
/Chris

_________________
LinHES R8.6.1
BE: AMD64X4, 4GB, Hauppauge usb tuners
FE1: Gigabyte F2A85X-UP4, nVidia GT640
FE2: Gigabyte GA-MA69GM-S2H, AthlonX2 4850E 2.5 GHz, 1GB, ASUS GEFORCE 7200GS 256MB
FE3: Asus Eeebox410