It must be Murphy's law.. starting a long easter weekend with disk problems on the scene.org server.. I just barely managed to log in remotely, and killed most services to stop the machine from hanging itself.
Running a badblock scan now, but it doesn't look very good..
Update 09-04-2004: Managed to get the website online on another server temporarily. The rest of the services will be offline until next tuesday i'm afraid.
Update 13-04-2004: Tonight I will try to duplicate the broken disk onto a new one with ghost, hope this is going to work...
Update 13-04-2004 (2): Hm.. ghost is running, but after the first few bad blocks the 'time remaining' counter went up to 24 hours.. so this might take some time..
Update 14-04-2004: Ok, ghosting has worked. We now have a duplicate of the broken disks. I can mount all the filesystems, but there are some problems in the filesystem which can only be corrected by running a command that can also result in total dataloss... so I don't think i'm going to try that. Real dataloss at the moment is not really that bad, nothing we can't put back from our ftpmirrors around the world. But I think we will have to buy a hardware raid array now and copy all files to that new volume. |