Date: Fri, 6 Oct 2000 10:21:19 -0700 (PDT) From: Anthony Fleisher <fleisher@mind.net> To: freebsd-questions@freebsd.org Subject: Request help with Filesystem Problems Message-ID: <Pine.BSF.4.21.0010060958190.72619-100000@takhus.mind.net>
next in thread | raw e-mail | index | archive | help
I have been having problems with one of my filesystems on a SCSI disk that seem to be related to disk failure, so am planning on moving the system from the SCSI disk to an IDE drive that will replace it, and thought I would drop a note to this list to see if I could get some advice/pointers from some people that may have already been through this process. System is FreeBSD 4.0-RELEASE The troubles: I am seeing lots of errors like this in messages and dmesg output: (da0:ahc0:0:0:0): READ(10). CDB: 28 0 0 7b 90 c0 0 0 10 0 (da0:ahc0:0:0:0): MEDIUM ERROR info:7b90c1 asc:11,0 (da0:ahc0:0:0:0): Unrecovered read error sks:80,9 (da0:ahc0:0:0:0): READ(10). CDB: 28 0 0 7b 90 c0 0 0 10 0 (da0:ahc0:0:0:0): MEDIUM ERROR info:7b90c1 asc:11,0 (da0:ahc0:0:0:0): Unrecovered read error sks:80,9 (da0:ahc0:0:0:0): READ(10). CDB: 28 0 0 7b 90 c0 0 0 10 0 (da0:ahc0:0:0:0): MEDIUM ERROR info:7b90c1 asc:11,0 (da0:ahc0:0:0:0): Unrecovered read error sks:80,9 The dump that I tried to do on the /usr/local filesystem last night looked like this (I use amanda to dump to another host): DUMP: Date of this level 0 dump: Fri Oct 6 01:04:46 2000 DUMP: Date of last level 0 dump: the epoch DUMP: Dumping /usr/local to standard output DUMP: bad sblock magic number DUMP: The ENTIRE dump is aborted. -- I have some concerns about my ability to get this system back up after the reboot necessary to install the new IDE disks, given these errors, so here is my plan: 1. use tar via ssh to make a copy of the entire filesystem to a remote system (I have enough diskspace) in case the partition does not want to come back online after the reboot. 2. shutdown the system, install 2 new IDE drives that have been had the partitioning and newfs done on another system. 3. bring the system back up, mount the new partitions under /mnt and use several tar pipelines to copy the files from the existing filesystem over to the new disk. (I will be moving / /var /usr and /usr/local). (If this doesn't work, use the fixit disk to bring up networking, copy the tarball from the remote system, and untar it onto the new disk.) 4. Change etc/fstab on the new drive to use the IDE rather than SCSI partitions. 5. Install bootblocks on the new drive. 6. run MAKEDEV in the /dev directory on the IDE disk to make sure that all my IDE slices/partitions are there. 6. Reboot off of the IDE (leaving the SCSI drive where it is, unreferenced). 7. At this point I expect the system to come up and work as before, only now using the IDE rather than SCSI. --- So, my questions: 1. Is there anything that I have forgotten in the above procedures? 2. Are there any problems that might arise? 3. I am expecting that, as long as the root filesystem comes back, the untarring of the backup created in step 1 above should keep all file permissions, in the event that the /usr/local system is unable to be mounted after the reboot. Can someone confirm this (assuming -p to tar)? 4. Are there any critical files that may have problems with tar? If you would like any further information, please ask. Thank you very much for any help. TOny. -- Anthony Fleisher <fleisher@mind.net> System Administrator InfoStructure Ashland, Oregon Voice: (541)482-8324 Fax: (541)488-7599 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0010060958190.72619-100000>