jimbly Posted May 12, 2006 Report Share Posted May 12, 2006 Hi there im new here to the forum we have problems with our backup 2Ghz G5 xserve 10.4.4 with Xserve Raid which has 4x 250 drives using scsi card ATTO UL3D connected to a Sony SDX-700C with LIB-81 tape library we have 2 backups running our daily backup goes every evening, backs up about 30-40gb of daily work. recycle backup our weekly backup goes every saturday morning. backs up about 450Gb onto 4-5 tapes previously was recycle backup, changed to normal incremental. every 2 weeks we change the tapes to a new set daily backup runs fine, no problems but every second saturday the weekly backup crashes, xserve reports a kernel panic, no crash log from retrospect. this is normally the weekend after we change the tapes, the following week it runs fine. this is puzzling us somewhat i suspect something to do with the scsi card, but it is so consistent, always second saturday. when the weekly tapes were on recycle backup, i tried deleting the tapes before, made no difference any ideas David --- Sun May 7 05:06:54 2006 panic(cpu 0 caller 0x002D57C4): IOGMD: not wired for getPhysicalSegment() Latest stack backtrace for cpu 0: Backtrace: 0x00095718 0x00095C30 0x0002683C 0x002D57C4 0x002D59E0 0x002D563C 0x00745D38 0x007460DC 0x0043C6F8 0x00448FCC 0x0044D3F4 0x0043F1FC 0x00458F4C 0x0043ED2C 0x0043F31C 0x0074348C 0x002EAB0C 0x0008CB88 0x000291C0 0x000233AC 0x000AC02C 0x59602A58 Kernel loadable modules in backtrace (with dependencies): com.apple.iokit.SCSITaskUserClient(1.4.4)@0x741000 dependency: com.apple.iokit.IOSCSIArchitectureModelFamily(1.4.4)@0x437000 dependency: com.apple.iokit.IOStorageFamily(1.5)@0x4fd000 com.apple.iokit.IOSCSIParallelFamily(1.4.0)@0x454000 dependency: com.apple.iokit.IOSCSIArchitectureModelFamily(1.4.4)@0x437000 com.apple.iokit.IOSCSIArchitectureModelFamily(1.4.4)@0x437000 Proceeding back via exception chain: Exception state (sv=0x232C1C80) PC=0x9000B208; MSR=0x0200F030; DAR=0x00323000; DSISR=0x42000000; LR=0x9000B15C; R1=0xBFFF9B70; XCP=0x00000030 (0xC00 - System call) Kernel version: Darwin Kernel Version 8.4.0: Tue Jan 3 18:22:10 PST 2006; root:xnu-792.6.56.obj~1/RELEASE_PPC --- --- + Normal backup using WEEKLY BACUP at 6/5/2006 08:30 To backup set Weekly C… - 6/5/2006 08:30:21: Copying PSP BACUP on RAID… 6/5/2006 19:33:14: Comparing PSP BACUP on RAID… ? Retrospect version 6.1.126 launched at 7/5/2006 05:06 Quit at 8/5/2006 15:55 ---- Link to comment Share on other sites More sharing options...
jimbly Posted May 12, 2006 Author Report Share Posted May 12, 2006 forgot to say, using retrospect 6.1 Link to comment Share on other sites More sharing options...
twickland Posted May 12, 2006 Report Share Posted May 12, 2006 It's not clear from your post: Did you have problems with your weekly backup when you used to perform a Recycle Backup, or did the problem only start with your New Media backups? Link to comment Share on other sites More sharing options...
jimbly Posted May 15, 2006 Author Report Share Posted May 15, 2006 we previously had the same problem with recycle backup. The backup went ok this weekend. we were thinking due to the amount of data we backup, would be better to change to normal incremental as each week it takes about 24hrs for the weekly backup to go through. Link to comment Share on other sites More sharing options...
jimbly Posted May 15, 2006 Author Report Share Posted May 15, 2006 small update. last weekend the crash happened during the comparing stage --- + Normal backup using WEEKLY BACUP at 6/5/2006 08:30 To backup set Weekly C… - 6/5/2006 08:30:21: Copying PSP BACUP on RAID… 6/5/2006 19:33:14: Comparing PSP BACUP on RAID… ? Retrospect version 6.1.126 launched at 7/5/2006 05:06 Quit at 8/5/2006 15:55 ---- this week because it is incremental, the backup only took 40mins for an 11Gb update has anyone else had problems with crashes during the compare stage? Link to comment Share on other sites More sharing options...
rhwalker Posted May 15, 2006 Report Share Posted May 15, 2006 Quote: has anyone else had problems with crashes during the compare stage? No, not in almost a year of use through the various updates of Retrospect 6.0 through 6.1.126, various RDU (none through current RDU 6.1.5.102), Mac OS X Server 10.4.2 through 10.4.6, Xserve G5 single 2.0 GHz 2GB RAM, ATTO UL4D driver 3.50 firmware v1.50, Exabyte VXA-2 1x10 1u. In fact, we've never seen a kernel panic. But then again, we aren't doing the amount of data that you are - our full backup is only about 70 GB, and daily incrementals are about 4 GB, but our Xserve is cranking away doing mail, DNS, and AFP at the same time. We did have some correctable ECC errors at first only when Retrospect was being run, but replacement of that memory module fixed that. Link to comment Share on other sites More sharing options...
ghostman Posted June 6, 2006 Report Share Posted June 6, 2006 We were having problems with two XServe G4s, OS X 10.4.4 Server, Atto UL3D cards, SDLT 600 drives, and Retrospect 6.126. Turns out it SCSI card placement is the key. Everything worked great once we moved the UL3D card to the same riser as the video card. The riser that came with an ethernet card has some issue with anything other than an ethernet card Link to comment Share on other sites More sharing options...
jimbly Posted June 28, 2006 Author Report Share Posted June 28, 2006 we upgraded to the LSILogic 1030 and updated the drivers, to see if it was the atto card, and it hasn't crashed. problem solved Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.