October 23, 201312 yr I am running 5.0 final It has been running great since the upgrade. example of log Oct 22 06:26:47 Tower kernel: Result: hostbyte=0x04 driverbyte=0x00 Oct 22 06:26:47 Tower kernel: sd 1:0:3:0: [sdh] CDB: Oct 22 06:26:47 Tower kernel: cdb[0]=0x28: 28 00 00 87 37 47 00 00 08 00 Oct 22 06:26:47 Tower kernel: end_request: I/O error, dev sdh, sector 8861511 Oct 22 06:26:47 Tower kernel: md: disk6 read error, sector=8861448 Oct 22 06:26:47 Tower kernel: sd 1:0:2:0: [sdg] Unhandled error code Oct 22 06:26:47 Tower kernel: sd 1:0:2:0: [sdg] Oct 22 06:26:47 Tower kernel: Result: hostbyte=0x04 driverbyte=0x00 Oct 22 06:26:47 Tower kernel: sd 1:0:2:0: [sdg] CDB: Oct 22 06:26:47 Tower kernel: cdb[0]=0x28: 28 00 00 87 37 47 00 00 08 00 Oct 22 06:26:47 Tower kernel: end_request: I/O error, dev sdg, sector 8861511 Oct 22 06:26:47 Tower shfs/user: shfs_readdir: readdir_r: /mnt/disk6/Movies (5) Input/output error Oct 22 06:26:47 Tower kernel: md: disk4 read error, sector=8861448 Oct 22 06:26:48 Tower kernel: program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO Oct 22 06:27:19 Tower last message repeated 221 times Oct 22 06:28:20 Tower last message repeated 410 times Oct 22 06:28:42 Tower last message repeated 163 times Oct 22 06:28:43 Tower kernel: sd 1:0:3:0: [sdh] Unhandled error code Oct 22 06:28:43 Tower kernel: sd 1:0:3:0: [sdh] Oct 22 06:28:43 Tower kernel: Result: hostbyte=0x04 driverbyte=0x00 Oct 22 06:28:43 Tower kernel: sd 1:0:3:0: [sdh] CDB: Oct 22 06:28:43 Tower kernel: cdb[0]=0x28: 28 00 00 87 37 47 00 00 08 00 Oct 22 06:28:43 Tower kernel: end_request: I/O error, dev sdh, sector 8861511 Oct 22 06:28:43 Tower kernel: md: disk6 read error, sector=8861448 Oct 22 06:28:43 Tower kernel: sd 1:0:2:0: [sdg] Unhandled error code Oct 22 06:28:43 Tower kernel: sd 1:0:2:0: [sdg] Oct 22 06:28:43 Tower kernel: Result: hostbyte=0x04 driverbyte=0x00 Oct 22 06:28:43 Tower kernel: sd 1:0:2:0: [sdg] CDB: Oct 22 06:28:43 Tower kernel: cdb[0]=0x28: 28 00 00 87 37 47 00 00 08 00 Oct 22 06:28:43 Tower kernel: end_request: I/O error, dev sdg, sector 8861511 Oct 22 06:28:43 Tower kernel: md: disk4 read error, sector=8861448 Oct 22 06:28:43 Tower kernel: REISERFS error (device md6): zam-7001 reiserfs_find_entry: io error Oct 22 06:28:43 Tower kernel: REISERFS (device md6): Remounting filesystem read-only Oct 22 06:28:43 Tower kernel: REISERFS error (device md6): zam-7001 reiserfs_find_entry: io error Oct 22 06:28:45 Tower kernel: program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO Oct 22 06:29:18 Tower last message repeated 228 times webpage still shows all disks are fine and green I got an email saying unraid was OK system still responds and except disk 6 and 4 which are empty John P syslog-2013-10-22.zip
October 23, 201312 yr I assume that disks 4 and 6 ate not meant to be empty If so it is likely that file system corruption is what is stopping files showing up. To correct the reiserFs errors you want to stop the array, start it again in maintenance mode, and then run reiserfsck from a telnet/console session against the problem drives. E.g. reiserfsck /dev/md6 Since you know you have corruption it is likely the above command will give an error at some point, but it should then suggest what is needed to fix the corruption. Typically this is to run reiserfsck with additional parameters to tell it to correct errors found. If you are not sure you should check back here in the forum before proceeding further. Note that on large drives a reiserfsck check can take hours so make sure you plan it for when you can afford to have the array off-line for that length of time.
October 23, 201312 yr Author Thank you for responding Smartrpt is failing on both disks with 67 errors each list. I will stop it Wednesday night and try it. john
October 23, 201312 yr Author More info here I think it is a power spliter that is flackey. My question is that UNRaid still thinks all the drives are ok. Why? None of them have red balled yet.
October 23, 201312 yr More info here I think it is a power spliter that is flackey. My question is that UNRaid still thinks all the drives are ok. Why? None of them have red balled yet. The errors have only been read errors. A write error will cause a disk to be taken out of service (red ball). unRAID main should show read error counts.
October 23, 201312 yr Author I plan on shutting down the system and checking cabling for power on those drives. rebooting and then doing a smart test on both drives. They both fail now. stop the array and do a reiserfs test to see what is happening. Very fustrating as the server has been running for 3 years now with a few problems. One drive dies but rebuilt wrong with lots of reiserfs errors. Some data loss. One drive died and then one fell out of the array, No data loss there. A few dead drives one at a time with normal rebuilds. Nobody every said this was cheap having a home server with lots of TB of storage. John
October 24, 201312 yr Author WOW I love UNRaid I shutdown checked cables and restarted. All is well now, no rebuild or hokus pokus needed. That makes 3 times I would have lost it all with raid 5 system. John
Archived
This topic is now archived and is closed to further replies.