Jump to content


  • Posts

  • Joined

  • Last visited

dgtldan's Achievements


Newbie (1/14)



  1. Would it make sense to run all the file servers, Mac and PCs that are being backed up, and the Retro backup machines on GMT? Has anyone tired this to eliminate this issue? TAI
  2. I have not seen crashes at launch, but have had issues with Retro hanging up in various operations phases on a DP G5 2GHz box. Sometimes it hangs suring a scan, other times i have had it hang after it completed a BU and the tape, AIT-2, retentioned itself. I am using a UL3D with the latest drivers from Atto. Ironically it worked great for about two weeks before any issues came up and no other software was added during that time. Any remedies? TIA
  3. I found out the storage sets are over a Tera Byte, which is causing the said behaviour. Isn't Retrospect suppose to warn you of that? Thanks.
  4. Hello, Recently I have noticed a couple of backup sets with members missing. It seems that Retrospect, 5.0.238 RDU 3.4.103 running on OS9.2.2, is actually marking them as missing, and the dates on the said missing tapes are set to Jan, 1 1904! How I found out was observing a backup script running very slowly and after stopping it manually I noticed the log said no files copied. Upon further investigation and looking at the members, I saw the missing members. They would have been the tapes that were being written into. The I looked at other backup sets and found a few more missing tapes. I can not make the system accept them as found set, nor can I verify them without verifying the whole set which can take days! Any ideas as to what is causing this? Thanks.
  5. We share the same issue. We observe this same issue on 3 entirely independent backup systems that are backing up multiple file servers over a network. The servers are a mix of Windows 2003, and SGIs. As far back as 4.2 on OS9, the response from Dantz was to make sure the clocks were set correctly, this is obviously not the answer. Every server, and Retrospect Mac station on the network, including the client that upload data to the servers are synchronized to the same NTP time server and their clocks are within seconds of eachother, clearely this is not a factor, or is it? Typical data on servers are about 6 TB and scripts are written for incremental backups. Our tape costs around this time of year and in 6 months from now, is just astronomical! Unfortunately it takes 6 months to see if any changes have worked! Any ways or ideas to resolve this? Thanks.
  6. Try the loader on ID 5 and keep both the loader and the drive on the same SCSI buss and LUN. I have done this on four models of Qualstar libraries with only one drive in the library, and they all work with 5.238 working under OS9. I have never tried the libraries with multiple drives using the Windows version. I have noticed that AIT-1 drive want to be set to ID 2, and AIT-2 drives want to be set to ID 4. I hard set the SCSI address on the drives with jumpers, but make the configuration:drive menu on the library have the same address as was set with the jumpers on the drive. Enjoy. --- Dan S.
  7. I have been able to get several Qualstars, even ones not in the support list, working by: Turn on SCSI Logging, Do a rescan from "Device Status" page, Review the Log for a line of two that reads someting like "Drive at [sCSI ID], Loader at [sCSI ID], Loader expected drive at [sCSI ID]". Change the SCSI ID to what the log reports it should be. I am not sure why this is... perhaps someone can explain this behaviour in detail. Best.... -- Dan S.
  8. Since my original post, i see that others have suggested this in alternate forums, having this added as a feature is a great tool. In our case, data is available on the servers often for many days, however, time is limited to perform all the backups and do a verify pass as well as a nightly routine. Hence for the feature ability of performing complete or partial verifys via a script. Thanks.
  9. Setting the Drive to ID 2 and Loader to ID 5 worked on a TLS-4220. The TLS now reports 1.31 under version in the Device Status page. Ironically this is the same firmware it was reporting before from Apple System Profiler. I had the Drive set to ID 0, guess that made RS unhappy! The only odd thing now is that i have to keep a tape in the drive or the library will not perform a scan media! Any reason for this to happen? I will test a TLS-4222 later and report the results as well. Would an X80 work this way as well, emulating a Exabyte 480? Thanks Amy.
  10. We are experiencing same issues with a slew of units we have tested. Various TLS series, 4410, 4410A, 4220, 4222, which are essentially exactly the same mechanism, do not show up in the devices page. The response from tech support was that it may be because they have the I/O port or Barcode reader fitted in them, however, I have other Qualstar libraries with I/O ports and Barcode readers fitted which work correctly. Quite honestly, we don't care for the proper operation of the I/O port or the Barcode reader as we well understand RS not dupporting these devices for the current Mac OSs. It sure will be good to get all the same family of libraries running though! Any support is appreciated.
  11. Hello, Which HVD SCSI card is supported for MAC OS9? Does Retrospect support the Sony SDX-510c, the HVD version of the SDX-500 AIT-2? Thanks.
  12. Hello, We recently upgraded from 4.3 to 5 Server and lost compatibility with a few of our loaders. Qualstar TLS-4220, TLS-4210 A, Exabyte X80 with native, 480 or, X200 emulation, can no longer be seen under the devices page in Retrospect and are reporting "no driver". All the above noted devices work with 1.8 ADK in 4.3. Is anyone experiencing similer issues? Ironically the reason we upgraded was because one of our TLS-4210 A devices stopped working under 4.3! Upgrading to 5 Server solved this issue, but introduced the above noted situation. All loaders have AIT-2 drives, run on OS 9.2, are on same type of G4s fitted with gigabit cards, and dual channel Atto UL2D SCSI cards. TIA
  13. I know I can make the verify as part of a backup script, but this extends the time of each backup script. I am hoping to be able to perform a full or a partial backup of a catalog during the day when the backup libraries are sitting idle. We backup large chunks of data overnight via multiple libraries running on Mac platforms. Does the Windows version support this type of scripting? Wouldn't it be more efficient if there was a way to verify full or partial scripts, partial being from the night before's snapshot, via a script? TIA
  14. After upgrade from 4.2 to 5.0 Server, i am noticing a significantly slow copy time of small size files, typically files under 100K. Larger file sizes, 1 meg or more, get coppied much faster. This was not apparent when i was using 4.2. Backups are done over a 100BT network of servers mounted on a G4 that is running 5.0 Server via Appletalk. The scripts and the storage sets are kept on the G4. Backup device is a SCSI X200 with AIT drive. Anyone else experiencing this? Any thoughs as to what could slow down the write cycle?
  • Create New...