Jump to content


  • Content count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About amcmis

  • Rank
    Occasional Forum Poster

Profile Information

  • Gender
    Not Telling
  1. amcmis

    Error Codes 102 And 205 During Restore

    Say, is anyone with Retrospect, like Robin, actually reading this forum these days? So I have put device logging up to 7, cleaned the drive, got a fresh tape, ran what would seem to be a good backup (no verification), then turned around to restore it. I have captured the logs for your reading pleasure. In this case it got to withing 30GB or so of a 200 GB restore, then froze up and the changer dropped out of sight, the drive still shows but reads as Not Responding. In the environment tab the Quantum drive still shows with (device busy) and the changer is listed with (unknown response) and no Driver. I rescan in the environment tab and the drive goes to (unknown response as well. We will have to power cycle now. Is there anything we can derive from these logs? LTO3-BackupGood.txt LTO3-RestoreErr.txt
  2. Please post the new link, and/or please fix your DNS. Edit: I still think there is funny with DNS, my internal Win2003 DNS servers are not able to resolve www.retrospect.com via the root hints. But forums.retrospect.com works fine. I poked around a little and see that there are a lot of CNAMEs out there, maybe there is a dynamic redirection as well? Anyway, I plugged in a forwarder for retrospect.com to and now I can at least see the HC database.
  3. amcmis

    Backup Failed After 507Gb. Can't Read File Error -1019

    This is a very interesting post to bump. I am wondering if this is related to a problem we have had for years, namely that grooming will fail on large Disk backups from time to time, requiring us to rebuild (not repair) the catalog and reattach them to all scripts afterwards. (Quite a pain, considering how long it takes to rebuild. And if you forget, no backups until you remember.) We have even taken to just resetting backups to quickly solve the problem, but that leaves us with no historical snapshots. I never took statistics on this, but the last time I was informed the catalog was corrupt, (about a week ago) the backup set was just about half a terabyte. I have never heard about a paged pool memory hack, etc. Can someone point me to any other info on this 'limit' and if it has been acknowledged by the powers that be? And yes, I'm still on 7.6, latest and latest hotfix.
  4. Since www.retrospect.com doesn't resolve anymore, where can I find the above device support DB? I posted a question on TL2000 w/LTO5 SAS, this post gives me hope it is actually supported. Time is an issue, as I need to make a purchase decision quite soon, and we may have to abandon Retrospect if it doesn't keep up with device support.
  5. amcmis

    Error Codes 102 And 205 During Restore

    Bump, I sayyyy Bump! Oh, and go look at my post in the hardware section on LTO5s while you're at it.
  6. Hoping this forum is going to pick up the pace soon. I am a long time Retro Multi 7.6.x user and have (mostly) been successfully using a Dell Powervault 124T Library w/ SCSI LTO3s (1 Quantum, 1 IBM) and one with a SAS LTO4. I am considering upgrading to LTO5 this year, in particular a Dell PV TL2000 SAS library w/single LTO5, unklnown make. I can only see 1 LTO5 (Tandberg) listed at all in hardware compatibility notes at the Rovi site. I realize that my 7.6x probably won't get the updates, but does 7.7.x have support for this device?
  7. amcmis

    Error Codes 102 And 205 During Restore

    I will add my voice to this. We have recently moved our tape auto-loader from one Dell server to another, using the same SCSI card, cable etc, and we are now seeing this failure on verification scans, causing the tape drive and autoloader to disappear from Retrospect. Only full power cycles bring it back. I have recently updated the autoloader AND drive firmware to latest. I did just update to the latest driver hotfix, but have no feeling that it will make a difference, as I have an absolutely identical setup at another site that does not have this issue. It may well be that the drive tself is having problems, but wanted to see if there is anything else I can get from these error messages. Dell PowerVault 124T (1.08) FW 85 Quantum Ultrium 3 DC (1.41) FW 2181 Moved Adaptec 39320A from PE 840 to a PE2950, using same cables. Retrospect 7.6.123 w/ Driver/Hotfix (was An interesting note: I just checked the last restart of Retro with the one immediately following my hotfix. There appears to be 'unknown Windows errors' in the previous restart. Maybe the hotfix actually did something? I'll check back with results as soon as I get them. + Retrospect version 7.6.123 Automatically launched at 1/17/2012 1:51 PM + Driver Update and Hot Fix, version MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserDefaults): failed, winerr 1021, error -1001 MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserChosenExecuteHandlers): failed, winerr 1021, error -1001 MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserDefaults): failed, winerr 1021, error -1001 MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserChosenExecuteHandlers): failed, winerr 1021, error -1001 MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserDefaults): failed, winerr 1021, error -1001 MapError: unknown Windows error 1,021 TReg::Create(Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers\UserChosenExecuteHandlers): failed, winerr 1021, error -1001 2:35:39 PM: Library "Dell PowerVault 124T, ID 3:6:1" could not clean drive "Quantum Ultrium 3 DC, ID 2:6:0", a cleaning tape was not found or its cleaning count was reached. It was used 13 hours since the previous cleaning request. 3:09:57 PM: Drive "Quantum Ultrium 3 DC, ID 2:6:0" was cleaned. 3:09:57 PM: Drive "Quantum Ultrium 3 DC, ID 2:6:0" was cleaned. Exit at 1/17/2012 3:17 PM + Retrospect version 7.6.123 Launched at 1/17/2012 3:17 PM + Driver Update and Hot Fix, version
  8. Update and apology! I decided last night to give it another try. This time I noticed that the latest Retro release (.123) included a driver update not available to the one earlier version we were running. When I found the release notes for the driver/hotfix, I saw that the PV124T SAS was explicitly mentioned. And lo, the changer now works as it should. Questions remaining: - We are still running a full backup on the SAS LTO4 and once done will compare speeds with the previous nights backup on the SCSI LTO3. As of right now the speed looks to be near the same. Is this a throughput limit from retrospect, or might it be something that can be tweaked? For full disclosure, I still have the Adaptec 39160 card in the server, could that be competing for resources? - A niggly question about Retrospect and changers: If a backup finishes a tape and needs a new blank tape, it will always start at the highest numbered slot and work backwards. I would think for organization you would want it the other way around. Is this changeable, or is it a function of the hardware? Thanks in advance, we're making progress!
  9. PLEASE, AN UPDATE ON SAS SUPPORT AND/OR WORKAROUNDS! Hmmm, I thought this would be a hotter topic. I will volunteer my problem: A Dell Powervault 124T 16 tape changer with a Quantum LTO4 SAS mechanism. The server is a Dell PowerEdge 840 with a Dell SAS 5/E controller. (PCIe x16) So, this is intended to be an upgrade from our older PV124T/LTO3 SCSI based unit. We need the increased speed and capacity at our primary site, the older unit will pass to a satellite location. We use a Disk-Disk-Tape method, putting locally stored RDB files directly to tape at night for maximum streaming. Result 1: The drive and changer are visible with in Retrospect, all looks OK. Result 2: With a tape in the drive, backing up RDB files is either exactly the same, or possibly slower than the older drive. We ran the test during the day, so I can't be too scientific about the results, but my gut tells me I'm in trouble now. Result3: The changer shows tapes and barcodes, put will not play with the drive at all. No drag and drop. I can use the drive's web GUI to move tapes around, making it a very expensive single LTO4 drive. We have reverted to the old drive for now. With EMC's rollout of Retro8 for Macintosh, I am led to believe there is a commitment to this product, so if it's not going to be years, please give an ETA for drivers for what I have to believe will be mainstream products. And... are there any workarounds to the performance issues for the SAS drive itself? I think I read something in this forum about registry weaks for the SCSI/SAS card drivers, but didn't really get any explicit guidance. Thanks in advance, and don't make me leave the Retrospect fold. I'm too old to recreate all these scripts!
  10. OK, that was it on the Terastation. I cleared all of the security checkboxes and it worked. Shall I blame the Samba implementation of these NAS boxes as the culprit? This probably doesn't address all of the issues in this thread, as server security info from Win boxes does need to be backed up. And I'm still not happy that Retro cannot backup Security changes without backing up the WHOLE file again, per the TechNote.
  11. Whoa, I've been away and had no idea this thread piqued everyone's interest so much! OK, here's what I just did, and it's worse than I thought originally. I am doing a backup of a NAS box, so it's a SMB mounted share. (Buffalo Terastation v1) My destination is a "Disk" backup, which happens to be a 134GB disk at this moment. (Yup, I know this can't work in the long run!) My backup set is set to groom via "Retrospect predefined policy." For the test, I ran the first normal backup at 4:00 yesterday, and it backed up maybe 50GB of files, compressing about to 35GB in the destination directory. The Terastation had 2 shares, one with 1181 files, the other with 77968. It took 2:39 to complete. All good. The scheduled backup was for 9:00 last night, so I went home with crossed fingers. This morning, I looked and found yes, the backup ran, showing as 'normal', but took 2:57 to complete. And guess what, my backup set has 4 sessions now, with 1181 and 77968 files re(tro)spectively listed in the nighttime one, identical to the first one. My destination directory is now at 70GB. With fingers now uncrossed, but having read the archive bit stuff in this thread, I run the third normal backup. Well, it's 1:24 into it and 26GB completed. I am assuming my backup container will be around 105GB by the end of the third backup. THIS IS A SERIOUS PROBLEM!!! 2 I might live with, but every backup? The only solution I can think of is to do a Recycle every night. It would probably be shorter. Is this happpening with Tape as well? That would be extremely rude. amcmis EDIT: I did check the Security section, which was set to back file and folder for Servers, not Workstations. Does Retro call a generic NAS box a Server? And besides, there is no AV or other processes outside of Retro touching these files, to my knowledge. (Unless Buffalo RAID5 implementation is to blame? XFS file system?)
  12. Yup, I missed the details. Win2003 Server Std 32bit all around, latest patches. too. My test only includes Retro Clients to a local drive, F:, in DISK mode as opposed to FILE mode. I haven't tested C: as a source to F: as a destination, since it would be silly. Eventually I back up all local disks to a LTO2-L drive, but haven't seen this behaviour yet there. (Doesn't mean it hasn't happened, just haven't seen it...) I was using no verification at the time. I will try a couple more tests tomorrow to see how often I can reproduce it. jth
  13. 'Allo, Just installed a fresh copy of 7.5 Multi, got the latest update. Installed the client on second server. Set up a script to backup second server to a 'Disk' backup on the primary server. Use the RUN menu to launch it and create first NORMAL backup. All goes well. (around 10GB) So now I use RUN again to do a NORMAL backup expecting it to be very small, just a few changed things on C:... But lo, another 10GB or so goes to the backup, which now, after compression, is 17GB or so. Confused, I RUN a third time, same way, and this time it does the correct thing, backing up just a handful of MBs. So what's up? The first NORMAL would act as a RECYCLE, I know. Is this a known bug about the second? I had this happen with both of the client servers I'm backing up in this way. (In case you are wondering, it's a D-D-T strategy so I can keep the tape drive streaming...) TIA! jth
  14. Hiya, we are thinking of upgrading our Retrospect 6.5 to 7.5, and are wondering what the functional difference is between single server and multi-server? If we get SS and set up client software on other 'servers' will it refuse to backup, or will it backup basic data like a desktop, but not server-specific security info? If the latter, that might be OK... Also, what is the criteria Retro uses to define a 'server?' Is a FC4 linux box a server or a workstation? Novell? In multi-server, is there a special server client piece? Thanks in advance! jth
  15. Hi, thanx for getting back to me. Yes, it's 6.1 as of earlier this week, hoping it would solve some issues, but the symptoms started at the earlier versions of both Retro and OS. Before continuing, I want to stress that I'm not sure it is Retro causing this, I am also suspicious of Norton AV. If people think Dantz made a mess of Retro for OSX, try NAV. We use the 3rd drive for various type of scratch storage, so yes, it's a file backup. I did see some comments about the system overhead for that, but I was initially more concerned about the RAID comments. Any specifics I should know? If the answer is a definitive "Don't use file backup!" I would have to look into an external firewire, I think. As far as immediate crisis was concerned, early this morning we did try to sequentially kill all Retro processes, and LiveUpdate finally. The specific windows did disappear from the GUI, but we weren't able to take control, force-quit anything, etc. Maybe it's an OS corruption thing? In any case, we needed to do a remote restart and inconvenience everyone. *sigh* And just in case, even though it's really an OS subject, does anybody know how to reload the console GUI without shutting the system down? Some kind of force-logout from a remote shell prompt? Just checkin'... TIA! jth