DrBoar Posted November 14, 2008 Report Share Posted November 14, 2008 Backupserver. Dual core G4 running 10.2.8 Server and Retrospect 6.1.230 Windows clients runing 6.5.131, 6.5.138, 7.0.112 Macintosh clients 6.1.130 Backup 1&2 on alternating days (exept sunday) Backup 1. To external hardisk no problem Backup 2. Filebackup to a LaCie 1TB external USB HD connected to a G4/400 running 10.4.11 The G4 locked up during backup once or twise after a couple of backups I got "Error-24201 (chunk checksum did'nt match)" Things tried to fix it. Rebuilt catalog- OK Verify backup- "bad backupset header found (0x436f6e74 at 3,230,580) 1 files were not verified" I have been running disktools to verify that the extenal HD is OK I also moved the external HD from a Dlink USB hub to be directly connected to the USB2 port. 1. Can it be fixed? 2. How to avoid things like this in the future? TIA Quote Link to comment Share on other sites More sharing options...
twickland Posted November 14, 2008 Report Share Posted November 14, 2008 If you were able to rebuild the catalog successfully, you should be OK for the moment. Files not in the new catalog will be backed up the next time. More of a concern is why you had the problem in the first place. A lot of users--myself included--have experienced problems with these big LaCie drives, whether due to power supply problems or other reasons. These units are typically multiple drives in a single enclosure in a RAID 0 configuration. Because of this, the whole "volume" is less reliable than a true single drive would be, as the failure of any of the physical sub-drives will effectively wipe out all your data. I wouldn't rely on Disk Tools to give you the whole story. Because Retrospect is a disk-intensive application, it can give you a better advance warning of impending disk failure. Since you have two alternating backup sets, you won't be completely if the creek if/when the LaCie drive fails, but if I were you, I would begin planning for its replacement. Quote Link to comment Share on other sites More sharing options...
DrBoar Posted November 17, 2008 Author Report Share Posted November 17, 2008 Now the backup was up and running again, for a while. After putting in some old 256 MB DIMMs it appears that the G4/400 tend to lock up. I now have put back the orignal 128 MB DIMMs to see if this was the culprit, or if it is the LaCie that does something and then lock the OS. This LaCie is not raided, the cabinet is so small there is no way of getting two 3.5" in it. A more general concern I have is that the disks are crammed into boxes with no cooling/ventilation while being on 24/7 and also have periods of quite intensive backup activity Quote Link to comment Share on other sites More sharing options...
DrBoar Posted December 3, 2008 Author Report Share Posted December 3, 2008 Verification does not look to good, any mean of fixing this? + Executing Verify at 2008-12-02 11.14 To backup set winter08-09… Backup set format inconsistency (5 at 2601919) Bad backup set header found (0x466f726b at 2,601,924). Backup set format inconsistency (5 at 2602284) Bad backup set header found (0x436f6e74 at 33,600,392). Bad backup set header found (0x436f6e74 at 67,788,476). Bad backup set header found (0x436f6e74 at 69,516,004). Bad backup set header found (0x436f6e74 at 111,580,204). Bad backup set header found (0x436f6e74 at 114,666,124). Bad backup set header found (0x436f6e74 at 115,603,440). Bad backup set header found (0x436f6e74 at 258,610,652). 11 files were not verified. 2008-12-02 20.15.28: 11 execution errors. Remaining: 11 files, zero KB Completed: 878955 files, 288.0 GB Performance: 544.9 MB/minute Duration: 09.01.02 Quote Link to comment Share on other sites More sharing options...
lstone19 Posted December 3, 2008 Report Share Posted December 3, 2008 Twickland earlier made comments about LaCie drive reliability. I have a number of LaCie drives and while I've never had an actual drive or enclosure die on me, I've replaced a number of power supplies. Does you drive have the 4-pin power connector (4 pins in a sqaure)? Those power supplies are sufficiently unreliable that I now stock extras at home ready to replace one as soon as I start seeing problems. The power supplies can be ordered from LaCie as well as other electronics sources. Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted December 3, 2008 Report Share Posted December 3, 2008 Backupserver. Dual core G4 running 10.2.8 Server and Retrospect 6.1.230 ... Backup 2. Filebackup to a LaCie 1TB external USB HD connected to a G4/400 running 10.4.11 Can you please explain this discrepancy? Is Retrospect running on a "Dual core G4" (no such beast) running (very old) OS X Server? Or is it running on a G4/400 running Tiger with USB hard drive attached? How exactly are things configured? Quote Link to comment Share on other sites More sharing options...
DrBoar Posted December 4, 2008 Author Report Share Posted December 4, 2008 To clarify the setup Room 1 Dual CPU G4 (1 GHz) running Retrospect 6.1.230 and the old 10.2.8 Server OS. The original setup was tapes with two sets for alternate weeks. This was replaced by hard disks also alternate weeks. To get rid of the hassle of swapping out the other weeks harddisks for safekeeping I decied to alternate between disk backup at the server and file backup to an other computer located somewere else. So enter room 2 with G4/400 running 10.4.11 and external USB disk Not running retrospect at all but sharing its USB disk. Setup Mo Disk backup to USB drives on the 10.2.8 server Tu File backup to shared usb volume on the 10.4.11 We Disk backup to USB drives on the 10.2.8 server Th File backup to shared usb volume on the 10.4.11 Fr Disk backup to USB drives on the 10.2.8 server Sa File backup to shared usb volume on the 10.4.11 I get backups that are on different physical locations that are of different types running on different OSs without having to remove disks for safe keeping (due to Retrospects inability to handle the dateformat in NTFS volumes I start a new set at every change to summer or winter time and the old sets are stored in a safe room) Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted December 4, 2008 Report Share Posted December 4, 2008 To try and condense the problem report: - Retrospect 6.1.230 (RDU unknown) running on a Power Macintosh G4 1.0 DP (either QuickSilver 2002, Mirrored Drive Door or FireWire 800 model), with OS X Server 10.2.8. - File Backup Set stored on locally connected USB2 hard drive (make/model unknown) works as expected. - File Backup Set stored on remote volume (file sharing protocol unknown) causes problems. That's really all we know with certainty. The G4 locked up during backup Which G4? The machine running Retrospect? Or the machine in the other room that happens to be acting as a NAS host for a shared volume? Neither of the two Power Macintosh G4 400 models that Apple shipped (AGP and Gigabit) included USB2 ports; how is the LaCie USB drive connected to the host machine? Quote Link to comment Share on other sites More sharing options...
DrBoar Posted December 8, 2008 Author Report Share Posted December 8, 2008 The G4 that locked up is the G4/400 that is a AGP sawtooth that use afp file sharing. (It has 1 GB RAM and a USB2 PCI card)(Sonnet Allegro USB2 I belive The Qucksilver 2002 has 1.5 GB RAM and also a USB2 card (Sonnet Allegro USB2 I belive) Retrospect 6.1.230 Device Access Version 1.0.107 Driver Update Version 6.1.15.100 Things are going downhill. I have tried to rebuild the catalog and I get (Trouble in Retrospect: Internal consistency check failed: Assertion check at "elem.c-923") Thanks a lot for putting up with my cryptic postings! Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted December 8, 2008 Report Share Posted December 8, 2008 The G4 that locked up is the G4/400 that is a AGP sawtooth that use afp file sharing You're approaching this as a Retrospect problem, when it sounds as if you actually have a problem with your file server hardware. All Retrospect is doing is writing files to the share; if the sever hosting that share is flaky, it can get overwhelmed by the effort and fail. If you have the option, I'd look at using a SATA host bus adapter in the Sawtooth with either an internal or external SATA drive. They're cheap, and faster then USB. I have tried to rebuild the catalog and I get (Trouble in Retrospect: Internal consistency check failed: Assertion check at "elem.c-923") There is no option to "rebuild the catalog" of a File Backup Set. The only choices are "Repair file backup set" and "Update existing catalog file." Which are you attempting? Since there is a probable issue with the system that is storing the File Backup Set, you should see how things behave after changing some of that system configuration. - What happens if you attempt to Repair the Backup Set or Update the Catalog file with the USB hard drive that houses it connected directly (via the USB card) directly to the machine running Retrospect? - What happens if you copy the File Backup Set ("winter08-09" and "winter08-09.cat") to another hard drive that is connected directly to the machine running Retrospect? Quote Link to comment Share on other sites More sharing options...
DrBoar Posted December 10, 2008 Author Report Share Posted December 10, 2008 I copied the file backup set to an other HD (Firewire) connected it directly to the Retrospect computer and used "Repair File backup set" I get the same error message as before after about 100 GB and Retrospect Quits. I do agree about the eSATA. However, with the next Retrospect version being intel only eSATA would exclude mac mini (that I was planning to use) as well as macbooks and iMacs. Leaving me with only Mac pro and Pro Book as options! It looks like I will buy an other 1 TB HD this time one with FireWire and make a new set. It seems I will get a Intel iMac from a leaving post doc so that will probably start to run Retrospect and then I will use the old dual 1 GHz as the afp server and then that flaky G4/400 will work on SETI or something less critical :banana: Quote Link to comment Share on other sites More sharing options...
DrBoar Posted January 13, 2009 Author Report Share Posted January 13, 2009 I treieed with a new nonIomega FW driwe. Backing up all stationary Macs over night was OK. During daytime backup of portable computers at least one uses seem to have pulled the network cable during backup. I got a network error and all following clients are not backed up and has this checksum error. :angry2: It would be nice to have safeguard. 1. Make temporary backup 2. Run verification 3.A. if OK append backup to backup file and catalog 3.B. If error abort backup of this client and then start with next client. It is not good if someones misshap will corrupt the backup for all other users Quote Link to comment Share on other sites More sharing options...
DrBoar Posted January 19, 2009 Author Report Share Posted January 19, 2009 I have a new backupset and a new checksum error. It started with a network error from a particular user. I asked him, and he had janked out the network cable while the laptop was about to turn of. After that I get only checksum error I tried Tools/repair/uppdate existing catalog file =internal consistentcy check failed I then try to Tools/repair/repair file backup set The snag is that the dialog ask for the catalog file, however the .cat is grayed out and I can only select the backup file??? Quote Link to comment Share on other sites More sharing options...
twickland Posted January 19, 2009 Report Share Posted January 19, 2009 The snag is that the dialog ask for the catalog file, however the .cat is grayed out and I can only select the backup file??? The language in that dialog box is confusing, but has long been that way (it seems to have been taken from the same dialog for opening a backup set catalog). Go ahead and select the data file. Quote Link to comment Share on other sites More sharing options...
DrBoar Posted January 20, 2009 Author Report Share Posted January 20, 2009 OK I have tried the repair. After about 50 och the 120 GB it gets stuck. The retrospect server is not doing anything but the file server is over its head. About 12 MB och 1024 MB RAM is free and two processes are stuck according to "top" in the termial. I did a frech restart of both computers before just to minimize such events. Could I try moving the external FWdisk from the file server to the Retrospect server and rebuild from there??? Then the traffic would be through the DMA of FW instead over TCP/IP TIA :happy: Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted January 20, 2009 Report Share Posted January 20, 2009 Could I try moving the external FWdisk from the file server to the Retrospect server and rebuild from there??? Didn't you say that you were doing exactly that? > I copied the file backup set to an other HD (Firewire) connected it > directly to the Retrospect computer Quote Link to comment Share on other sites More sharing options...
DrBoar Posted January 21, 2009 Author Report Share Posted January 21, 2009 (edited) No, not really. The volume for file backup is physically located in an other building and accessed through apple file share. So I used this mode of access also when trying to rebuild the catalog. Now I tried the rebuild with the Iomega Mac HD connected directly to the Retrospect server. It willl not mount! There is something in Iomegas formatting that stop if from mouning in 10.2.8 (Retrospect server) or an other older PPC running 10.3.9. It monts nicly in the G4 file server running 10.4.11 as well as a intel iMac also running 10.4.11 However, if the retrospect server access the the disk through fileshare over the network it does mount without any problem... So I did not see that one coming :confused2: Edited January 21, 2009 by Guest 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.