>>> I don't think it is the power supply. Running bacula causes the crash. >>> The machine only crashes when bacula is involved. This machine does not >>> do 24x7 backups. Full backups, for 3 clients, take at most 6 hours. >>> Incremental backups for these same 3 clients takes about 20 minutes. >>> Bacula may not be the root cause of the problem, but it is definitely >>> indirectly the cause of the problem. >>> >> >> *Nothing* should be able to cause the kernel to crash. It is either a >> kernel bug (not bacula's fault), or faulty hardware (also not bacula's >> fault). >> > > Yes, only faulty hardware or some code running with kernel mode > privileges, such as a kernel module or built-in driver for some piece of > hardware, can cause the crash. An easy way to test this is to backup to > another storage device. Since he is, I believe, backing up to a USB > connected hard drive, I suggest trying to back up to a SCSI or SATA > connected hard drive. If that works, then there is a bug in the USB > driver, a faulty USB port, or the USB drive itself is faulty. Code > running with user mode privileges, like Bacula, cannot possibly cause > the kernel to crash like this. If it can, then even that is a kernel bug. >
I don't think there is a problem with core components, CPU, memory, or power supply. All have been sufficiently pushed with various tests: cpuburn-in, mprime, memtest, etc. I doubt there is a physical problem with the external USB hard drive or the USB port itself. Although, I have frequently questioned the USB driver and kernel. Unfortunately at this stage of the game rolling another version of the kernel is not a trivial task. I last tried it with a 2.6.16.5 and it was far from smooth sailing. I reverted back to 2.6.13.4 as I did not have time to go through and correct all the programs that broke with the new kernel. I may end up having to bite the bullet in the future if the crashing becomes more frequent. I wish I knew for sure what the root cause is so I could put my focus there. Unfortunately, all I have so far is that the crash always happens when bacula is running. It has even happened during normal hours when a manual job has been run. As the external USB disk is rarely used except for backups there is some reason to suspect that part of the architecture. The only issue there is that outside of bacula I have formatted the USB disk several times, ran disk checks, and have written a bunch a files without any problems. Oh well, I guess this thread is out there in case this does happen to someone else besides me. Thanks. ------------------------------------------------------------------------- SF.Net email is sponsored by: Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
