Jump to content


  • Content count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About Cpuroast

  • Rank
  1. Yesterday I spent time manually removing ~200GB of old useless snapshots in 1 of our sets(1 at a time by clicking forget each time) and then launched the groom to delete the actual data when I left work. I come in the next day, the groom went fine ~230GB groomed, no errors, so I go back in the set and remove the remaining snapshops I wanted to remove, launch another groom, again all is fine. So I leave it at that ~250GB free on the volume, all is good and ready for the 12:00pm start ouur daily backups of the offices. So the 2 scripts execute at 12pm each in their own execution unit, 1 backing up pcs to Set-A-PCs and 1 to backup macs Set-A-Macs. They both store the backed up data to the same physical Set A which as I said is 4x ~300GB drives spanned. The Mac script launches without a hitch and completes without errors. The PC script stops right at the beginning with a now very familiar error: + Normal backup using Windows at 6/5/2009 12:00 PM (Execution unit 2) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) E-mail notification failed: error -1201 (stopped by operator) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) 6/5/2009 12:31:38 PM: Execution stopped by operator Total duration: 00:31:31 These spanned hard drives are used to backup 3 backup scripts, servers that run at 4am and the pc and mac scripts that run at 12:00pm. Why do 2 of these always run fine(mac and servers) and the 3rd one PCs doesn't to the exact same hardware? If it was the hard drives, the other scripts would fail as well, wouldn't they?The PC data has been re-catalogged several times over the years, so it can't be the catalog. Would grooming be the issue? I'm getting a little frustrated here.
  2. It's 4 drives spanned, but we have 3 of these sets, which we rotate weekly. Set A is 4 ~300GB drives spanned Set B is 4 ~300GB drives spanned Set C is 1 1.5TB drive on it's own Set C was added recently to extend the amount of data backed up.
  3. If it was a problem with the Device itself, wouldn't the other 2 Backup sets which are backing up to the same device just in different folders, have problems as well?
  4. So yesterday I swap in this week's backup set. Our 2 backup scripts execute, 1 of them is fine and fully completes without error. The other one starts spitting out the following errors + Normal backup using Windows at 6/1/2009 12:00 PM (Execution unit 2) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) E-mail notification failed: error -1201 (stopped by operator) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) 6/1/2009 12:47:41 PM: Execution stopped by operator Total duration: 00:47:31 The one that worked and the one that failed are both backing up to the same physical drive. So anyways, I kill the script, leaving the other one running which completes fine. I decide to try a re-catalog to try to fix it but end up with the following errors: + Executing Recatalog at 6/1/2009 2:13 PM (Execution unit 2) To Backup Set Backup Set A... Device trouble: "1-Backup Set A", error -105 (unexpected end of data) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) E-mail notification failed: error -1201 (stopped by operator) Device trouble: "1-Backup Set A", error -105 (unexpected end of data) 6/1/2009 2:15:03 PM: Execution stopped by operator Duration: 00:01:36 So again I go in the set's folder and delete the last day of rdb backup files it had(~35GB) and initiate a re-catalog of "1-Backup Set A" which completed 4 hours later. + Executing Recatalog at 6/1/2009 2:22 PM (Execution unit 2) To Backup Set Backup Set A... The locked preference on the Snapshots is not preserved after a recatalog. Please review the Snapshots in the Backup Set and make appropriate changes. 6/1/2009 6:28:45 PM: Execution completed successfully Completed: 450568 files, 380.1 GB Performance: 1961.9 MB/minute Duration: 04:05:49 (00:47:46 idle/loading/preparing) This is what the retrospect engineer I spoke to a few weeks ago suggested. So I re-configure the scripts to re-integrate the now I hope, fixed backup set, since all scripts forget about a set every time the catalog has to be re-created. Backup starts normally, and after about 1hour into the script with around 95GB of space still left free on the drive I get this: - 6/2/2009 2:17:25 PM: Copying user on User's Computer - 6/2/2009 2:19:31 PM: Verifying user on User's Computer Device trouble: "1-Backup Set A", error -105 (unexpected end of data) 6/2/2009 7:43:41 PM: Grooming Backup Set Backup Set A... 6/2/2009 7:59:17 PM: Groomed 3.3 GB from Backup Set Backup Set A. - 6/2/2009 7:59:19 PM: Copying user on User's Computer 6/2/2009 8:02:55 PM: Snapshot stored, 2,516 KB 6/2/2009 8:03:03 PM: 1 execution errors Completed: 49 files, 1.5 GB, with 12% compression Performance: 391.8 MB/minute Duration: 00:21:30 (00:17:36 idle/loading/preparing) When it happened Retrospect decides to automatically groom the set for some odd reason, even if it still had free space, and after about 5 hours it succeeded in grooming a tiny 3.3GB and continues with the backup cycle which is useless now as most people have turned off their computers, which means today's backup window was missed and it will have to go to tomorrow. No events pertaining to drive errors in the Event log are present, which leads to me believe that something is wrong with Retrospect. What can I do? Error -105 (unexpected end of data) has happened on 2 individual backup sets so far. It has never happened in the past with version 7, 7.5 or 7.6 when we were below the 1TB threshold. No one answered my question about the 1TB retrospect volume size limit since I posted it on May 15th. Thanks,
  5. Hi, We are using Retrospect 7.6.123 with hotfix We have 76 clients divided up into 3 logical backup sets which get backed up into 2 physical backup sets that we rotate every week. The physical sets consist of 4 hard drives configured in a single Server 2003 dynamic volume. We keep 3 snapshots and the active physical backup set gets groomed every Friday. A little while ago we were running out of space, so we added another 250GB drive to each of our physical sets using Windows Server 2003 Dynamic volume expansion. This brought the total capacity of each set to 1.08TB. We then started to run into -105(unexpected end of data) errors. I called Retrospect support and they suggested I simply erase the rdb files after the last successful backup and re-catalog. So I did on both physical sets, this seemed to fix it. Until just now, with ~200GB free space, I get another error -105 and it complains of lack of free space. So this got me thinking, errors only started happening when we finally broke 1TB in capacity on the volumes. Does Retrospect 7.6 have a maximum volume size limit it can cope with or can it simply uses as big a volume as NTFS does. Thanks,
  6. I disagree. I would very much like to see if the task is expected to finish in 5 minutes' date=' 5 hours or 5 days. If it's 5 minute, I get myself a cup of tea. If it's 5 hours, I come back later in the day or maybe next morning.[/quote'] I guess, an ETA or progress bar would be nice. I was just saying I'd rather have multiple snapshot removal without any ETA or progress bar than none at all.
  7. It's very sad to see required functionality take so long to implement. In my opinion, this functionality is crucial for the next release. I don't see how hard it would be to implement. Allow user to selected multiple snapshots via shift-click or ctrl-click, when clicking on forget, put all snapshots to be removed in a queue and then process them 1 by 1. No need even for a progress indicator or an ETA on when the task will be completed.
  8. How about, it's on the critical feature list, that should have been there from the start? Would you have an ETA on when this feature will be available. And I hope it will be implemented in a free upgrade say 7.7. We backup around 80+ computers. That get moved, renamed, user name changes. So many old snapshot ghosts that fallout of the grooming schedule, get left on the backup set. And we then have to manually forget the ghost snapshots manually when the backup drive gets full. Oh and we have 4 backup sets with catalogs at around 1.5GB each for around 900GB of data. Retrospect is supposedly a professional backup solution and it costs as much as one. We own an unlimited multi-server license with the open file backup option. And it wasn't cheap. I expect to see this necessity(no, it's not a feature) implemented ASAP, or we will start looking elsewhere.
  9. Hi, I just did what you suggested, shutdown restrospect, uninstalled 7.6. Installed 7.5.508, went to clients, chose update clients, chose the 6.1 Mac OS X RCU. It goes through the process and none of the clients have reverted. In the Operation logs I see update errors related to error -541 (backup client not installed or not running) with ncmxOpenDLFile: can't open "retropds_macosx_uni.pdu", error -1101 (file/directory not found) following that. And quite a few error -1101 (file/directory not found) aswell. I also tried installing the latest hotfix after that, and tried again, no change. Sounds alot like I'm going to have to manually uninstall and re-install all clients with the 6.1 version.
  10. Are you guy serious? This was a major screwup, and there isn't a fixed RCU to fix the clients from 7.6? We already purchased and implemented Exchange 2007 backup support, which isn't supported under Retrospect 7.5. Therefore reverting to 7.5 isn't possible.
  11. The Server is 7.6 upgraded from 7.5. The Client RCU I was trying to use to downgrade the clients is the 6.1.130.
  12. I just did that, and it started quickly going through the 19 clients, and it said : 0 updated sucessfully, 19 not updated. They are still running 6.2.229. I check the operation logs and get a bunch of update failures related to "error -523 (service transaction error)" and "error -541 (backup client not installed or not running)"
  13. What I don't understand is this feature has most likely been requested since restrospect 2.0, we are now at restrospect 7.6 and it's still not there. And now with the Mac 6.2.229 RCU screw up, and the fact that 19 of the 19 mac clients I updated via the RCU will have to be re-installed with 6.1.130 since 6.2.229 always turns off spontaneously, even after a "re-install, forget re-add" and re-added to the Retrospect Client Database. Which after the backup cycle runs will create 19x3 orphaned duplicate snapshopts x2 backup sets = 114 orphaned duplicate snapshots. Because Retrospect isn't smart enough to groom orphaned snapshots caused by clients re-adds, even though they have the exact same name, same IP and same password. And since you can only do 1 at a time, that means 114 x 15minutes = 1710minutes, that's 28.5hours of time wasted, with me having to sit there and do each of them manually, 1 by 1. Imagine if we had thousands of broken 6.2.229 clients. What hell this would? >If you need to correct the problem from a central computer you can download Retrospect 6.1.138 for Macintosh at http://download.dantz.com/archives/Retro-EN-6_1_138.dmg >Install 6.1.138 to a new directory. In Retrospect go to Configure>Clients and select Update All from the client menu. Select the Mac_X_Client_Update_6_1.rcu file from the OS X Client Updater folder. This process will downgrade the client to the prior version. The previous suggestion is useless, as the clients have to be on and functional for the downgrade to work. The fact that 6.2.229 spontaneously turns off doesn't help at all. Retrospect is supposed to be a good product. And it's not like the Retrospect Multi-Server with Open-Backup License is cheap. This should not happen Is there any testing and quality control done in the development process at all? I will be vary weary of mac client upgrades in the future. And I may even completely forget about the mac support and find another backup solution. Retrospect 7.5 now 7.6 might just be the last version of Retrospect we purchase.