adrianlambert Posted October 8, 2010 Report Share Posted October 8, 2010 We're currently running 8.2.0(399). A script runs at the end of the day that backs up data from a samba share to a afp NAS HD. Recently it has been stopping before the scan is complete (about 8 hours - 14TB) with the error "Execution stopped by operator" It was stopping at 5am. Obviously no-one (only two of us have access to retrospect) was in at 5am stopping the script. To make sure i set the script to start and hour and a half earlier, it stopped at 3:30am with the same error at, I'm assuming, the same point in the scanning process. This has never happened before the other day. Any ideas? Here's the log. I read in the KB that oserr 22 was of no importance. Can't find a reference to -1002 to establish if it's of value + Copy using gotitnightlybac at 7/10/10 (Activity Thread 1) MapError: unknown Mac error 22 TFolderLoc::StartFind: UFindFirstFile failed, /Volumes/gotitnightlybac, oserr 22, error -1002 To volume gotitnightlybac... - 7/10/10 7:30:00 PM: Copying Got_It_Storage ****** Warning: Copying Time Machine data is not supported. ****** MapError: unknown Mac error 22 TFolderLoc::StartFind: UFindFirstFile failed, /Volumes/gotitnightlybac, oserr 22, error -1002 MapError: unknown Mac error 22 TFolderLoc::StartFind: UFindFirstFile failed, /Volumes/gotitnightlybac, oserr 22, error -1002 MapError: unknown Mac error 22 Scanning incomplete, error -1002 ( unknown Mac OS error) 8/10/10 3:25:13 AM: Execution stopped by operator Quote Link to comment Share on other sites More sharing options...
Daniels Posted October 11, 2010 Report Share Posted October 11, 2010 Do you have Time Machine running on the computer and if so is it on a schedule? I have seen Retrospect report "Execution stopped by operator" if it hits a setting that causes the backup to stop such as hitting the end time, etc. Quote Link to comment Share on other sites More sharing options...
adrianlambert Posted October 12, 2010 Author Report Share Posted October 12, 2010 No, no time machine and no reason that I'm aware of. This Xserve's sole purpose is to run retrospect. Quote Link to comment Share on other sites More sharing options...
Daniels Posted October 12, 2010 Report Share Posted October 12, 2010 Is the Xserve set to go to sleep or is it always on because it sounds like it is losing connections to the shares and thus causing Retrospect to stop the execution and report the message? Quote Link to comment Share on other sites More sharing options...
Maser Posted October 12, 2010 Report Share Posted October 12, 2010 Because it sounds like it's stopping on scanning the drive at the same point now, you might have some new funky file/folder that Retrospect is barfing on and you'll have to isolate this. The easiest (?) way of isolating this would be to set up a series of "Favorite Folders" and just keep scanning those until you find the "bad" folder, etc... Quote Link to comment Share on other sites More sharing options...
rhwalker Posted October 13, 2010 Report Share Posted October 13, 2010 Because it sounds like it's stopping on scanning the drive at the same point now, you might have some new funky file/folder that Retrospect is barfing on and you'll have to isolate this. Yea, probably an engine barf is treated like a SIGTERM or SIGQUIT. See signal.h for details. Russ Quote Link to comment Share on other sites More sharing options...
adrianlambert Posted October 13, 2010 Author Report Share Posted October 13, 2010 (edited) Thanks for your input. We may never know. I've got it going again by recreating the user account on the NAS drive that receives the backups and alls fine. FYI we have lost a catalog this weekend and retrospect won't rebuild it, also the other 2 catalogs seem to have been kind of reset so the incremental backup wants to start from the start of 12TB so this may be the final straw. Trialling BRU server seeing as we have nothing left to lose with Retrospect. Edited October 13, 2010 by Guest Quote Link to comment Share on other sites More sharing options...
Maser Posted October 13, 2010 Report Share Posted October 13, 2010 Unfortunately, there seem to be a number of posts about users with NAS drives having issues that people with other types of drives as backup media are *not* having. But, also unfortunately, nobody has been able to post anything to any of the forums reproducing a specific set of problems with specific NAS drives (formatting, methods of connection/disconnection what have you...) I'm sure if somebody could give a specific reproducible case, the Engineers would be all over this... Quote Link to comment Share on other sites More sharing options...
Maser Posted October 13, 2010 Report Share Posted October 13, 2010 Thanks for your input. We may never know. I've got it going again by recreating the user account on the NAS drive that receives the backups and alls fine. FYI we have lost a catalog this weekend and retrospect won't rebuild it, also the other 2 catalogs seem to have been kind of reset so the incremental backup wants to start from the start of 12TB so this may be the final straw. Trialling BRU server seeing as we have nothing left to lose with Retrospect. Depending on how you "rebuilt" the user account -- backing up all the data all over again may be expected. If the extended attributes/metadata of the files are now *different* -- Retrospect will see them as different files and back them up again. I have no idea if BRU does the same level of file comparison for matching purposes or not. One would think so (at least I'd hope so), so things may not be any different with another product. Quote Link to comment Share on other sites More sharing options...
adrianlambert Posted October 21, 2010 Author Report Share Posted October 21, 2010 Sorry, by rebuilt, I meant that I recreated the user account on the NAS. The source share has remained a constant. I have also now tried using the NAS admin user account, this didn’t make any difference. In response to the above post about a lack of info, i'll try and address that. The NAS is a lacie Mini Ethernet Disk http://www.lacie.com/support/support_manifest.htm?id=10258 It is connected via Ethernet on a subnet x.x.1.x directly into an xserve running retrospect 8.2.0.399. Retrospect connects to the share on the Lacie via the NAS's built in afp service. (I have no details on this) The rest of the network is connected to the xserve via a separate Ethernet port using subnet x.x.0.x As of writing this I've also dug a little deeper and found afpd errors on the NAS log, so retrospect may not be the culprit after all, but to clarify, the backup succeeded on the first run after recreating the user account on the NAS but fails on each attempt following that. Just been out to buy a couple of cheapo USB HDs to replace the NAS, so probably won't have any more to add here. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.