arndwetzel Posted August 29, 2006 Report Share Posted August 29, 2006 MacPro Xeon Server 2.66 latest Apple Software Software Edition (Universal) with Retrospect 6.1.126 crashes erery time, either on manual or automatic backups. no matter if duplication or backup to file or write to a AIT2 Turbo. Anyone experienced same problems with Xeon Macs??? thanks for ideas Link to comment Share on other sites More sharing options...
Mayoff Posted August 29, 2006 Report Share Posted August 29, 2006 Can you give us a little more details on this: 1) When does the crash happen? During launch, during scanning, during matching, while writing to the media, during compare, during a restore? 2) What do you see during this "Crash"? Does the computer freeze? Does Retrospect report an error message? Any errors in the console? Link to comment Share on other sites More sharing options...
arndwetzel Posted August 29, 2006 Author Report Share Posted August 29, 2006 it crashes when ever retrospect starts exicution somtimes on preperation but definitly when adressing a target volume (checked it with each harddrive - 2x 250gb hitashi 2x 500gb maxtor) changed ram, changed hds, with a different machine, pretty much everyting. tried with regular system no server. console is empty, only reporting the sudden fail an programm crash. it may have to do with the sata2 driver in the system. tape drive is recognised, and eareses tapes but if its startet it imediately crashes without any reply of the system error handler. the are also no hint in the console concerning a ram-architectur fail i'm up to the point prossesor incompatability and/or sata2 thanx Link to comment Share on other sites More sharing options...
nekr0phage Posted August 30, 2006 Report Share Posted August 30, 2006 Hi, Can you install the Retrospect client on another machine and try backing up from there? That might be enough to determine whether the problem is with the SATA2 controller or not. Have you tried backing up subvolumes instead of the entire drive? Link to comment Share on other sites More sharing options...
waltr Posted August 30, 2006 Report Share Posted August 30, 2006 hi arnd, Quote: tape drive is recognised, and eareses tapes but if its startet it imediately crashes without any reply of the system error handler. is this a SCSI device? if so, what SCSI card are you using? Link to comment Share on other sites More sharing options...
arndwetzel Posted August 31, 2006 Author Report Share Posted August 31, 2006 hi waltr and foster, its a fw AIT2 Turbo drive - with subvolumes is the same - no duplication nor writing files; no network backups possible from clients. backing up from an other maschine ist not possible, cause the server ist behind a firewall in a dmz...so i havn't tested it, yet. same server version works fine on G4/G5 maschines even running osx 10.4.7 - it all seems to be a xeon issue. witch is even harder - i was setting back the sata2 hds to sata 1,5gb/s - same crashes no massage... i'm stuck - and i need a running solution thanx for assisstance arnd Link to comment Share on other sites More sharing options...
Mayoff Posted August 31, 2006 Report Share Posted August 31, 2006 Have you tried this MacPro with 10.4.7 workstation instead of the Server version to see what happens? We have not yet had a chance to test the MacPro Xeon at EMC Insignia. We are working to get the hardware as quickly as possible. We hope to do some testing with 10.4.7 Server later today to see if the operating system is causing any of these issues. Link to comment Share on other sites More sharing options...
rhwalker Posted August 31, 2006 Report Share Posted August 31, 2006 Quote: Have you tried this MacPro with 10.4.7 workstation instead of the Server version to see what happens? We have not yet had a chance to test the MacPro Xeon at EMC Insignia. We are working to get the hardware as quickly as possible. We hope to do some testing with 10.4.7 Server later today to see if the operating system is causing any of these issues. Robin, Be aware (if you aren't already) that 10.4.7 Server (universal) is a different animal than 10.4.x with 10.4.7 upgrade. Odd but true. Russ Link to comment Share on other sites More sharing options...
Mayoff Posted August 31, 2006 Report Share Posted August 31, 2006 It does look like we reproduced the problem with 10.4.7 Server (Universal) on an iMac. We are investigating it right now. Link to comment Share on other sites More sharing options...
timppas Posted September 4, 2006 Report Share Posted September 4, 2006 Hello from finland.. I have exactly same problem. did you find any solution? t:timo Link to comment Share on other sites More sharing options...
Mayoff Posted September 4, 2006 Report Share Posted September 4, 2006 To be notified when a solution is found, join this mailing list: http://list.dantz.com/mailman/listinfo/mac_intel_support We know exactly where the failure is during the backup but we need to investigate the cause a little deeper. I can say it is for sure isolated to 10.4.7 Server for Intel and does not happen on 10.4.7 server for PPC. Link to comment Share on other sites More sharing options...
arndwetzel Posted September 5, 2006 Author Report Share Posted September 5, 2006 hi all, thanks for your sherlok-work...I hope you find a quick way to fix it...I had to bring the maschine online...using superduper for the mirroring...the tape drive is off...not very professional in the eyes of my costumer... best incremental greatings arnd Link to comment Share on other sites More sharing options...
Mayoff Posted September 5, 2006 Report Share Posted September 5, 2006 I heard a rumor that 10.4.7 Server for Intel does not have rosetta support. That sounded crazy to me. Does anyone know for sure? Link to comment Share on other sites More sharing options...
arndwetzel Posted September 6, 2006 Author Report Share Posted September 6, 2006 hi mayoff, must have rosetta support, otherwise the installer wouldn't have started...an the same problem appears under the nativ OS for MacPro... ...and the binary server 10.4.7 works fine on G4 and G5 maschines using the same version of retrospect... solong arnd Link to comment Share on other sites More sharing options...
phil_w Posted September 8, 2006 Report Share Posted September 8, 2006 Its an issue with the Seagate drives that come in the Mac Pro (or any Seagates for that matter). I had to fix it by changing them for Maxtors and a couple of Hitachis although I think the latest driver update fixes the issue too. Link to comment Share on other sites More sharing options...
Mayoff Posted September 8, 2006 Report Share Posted September 8, 2006 Based on our research, I don't think it is a problem with the brand of hard drive. We have a bug open with Apple and we are working with them on an investigation. Link to comment Share on other sites More sharing options...
phil_w Posted September 9, 2006 Report Share Posted September 9, 2006 Sorry to hear that you still have problems - switching drive brands cured the issue immediately for me. I even cloned the existing 10.4.7 server and Retrospect from original drive to new (i.e. not a fresh install) and all worked OK. I agree its probably not the brand but probably more to to with the SATA controllers on the MacPro boards. Interesting to note also that my original setup contained 2 500 GB Seagates mirrored together and performance was SHOCKING for data transfer/copying. The 2 Hitachis that replaced them are fine. Link to comment Share on other sites More sharing options...
CallMeDave Posted September 9, 2006 Report Share Posted September 9, 2006 Quote: I agree its probably not the brand but probably more to to with the SATA controllers on the MacPro boards. OK, this is important. Mayoff and his team have been able to recreate a problem in an Intel iMac. I'd assume that the same iMac can successfully launch Retrospect under some other flavors of OS X, but fails with a specific OS (Mac OS X Server 10.4.7 (Universal)). If Phil is successfully running Retrospect under Mac OS X Server 10.4.7 (Universal), no matter what hardware, this would be important to confirm. - Phil, what is the build number of the OS that you're using? - Mayoff, how about the build number of the test iMac at EMCInsignia? - And is the iMac booting of its internal SATA drive? Or is it booting off an external FW drive? Link to comment Share on other sites More sharing options...
Mayoff Posted September 10, 2006 Report Share Posted September 10, 2006 We have not tried booting from an external drive, that isn't a bad idea. We are testing with the latest 10.4.7 universal release build downloaded from Apple. A very specific file system read API is failing during the start of the backup, so we know the exact nature of the failure and are consulting with Apple developer support. Those exact commands work with all other versions of 10.4.7 Server PPC and Workstation (PPC and Universal) I know that we tested with an Intel iMac but I don't know which HD is inside. My home Intel iMac has a Hitachi drive. Link to comment Share on other sites More sharing options...
rhwalker Posted September 11, 2006 Report Share Posted September 11, 2006 Just to toss out one more possibility, switching drive brands (as Phil Welbourn did) would have caused a new driver to be installed on the new disk, and also would have caused a new install to have been done on the new disk. Was this procedure (initialize disk, install driver, reinstall) tried on one of the failing systems? Also note that Arnd is using a firewire drive ("fw AIT2 Turbo drive") not SATA. Perhaps the FW driver has issues that the other drivers do not. Just a thought for a regression test. Russ Link to comment Share on other sites More sharing options...
phil_w Posted September 11, 2006 Report Share Posted September 11, 2006 Build is 8K1079. Interestingly, during the nasty bit (when nothing worked), a scan of an external drive (firewire or usb memory stick for example) Retrospect was fine. Only when scanning the internal drives would it crash. Hope this helps. The 'fixed' MacPro still going strong - pumping data to a SCSI Ultrium 3 at great speeds. Also - no amount of initializing, OS X server reinstalls cured the problem. Only the drive swap. None of it makes an awful lot of sense, I know. Wouldn't normally dive straight in there with new kit as a server (would normally wait a while) but my customer was desperate and we couldn't get hold of a G5 in a hurry. Link to comment Share on other sites More sharing options...
Mayoff Posted September 11, 2006 Report Share Posted September 11, 2006 Phil, I am passing all of your feedback to engineering. These are very interesting observations on your part. Link to comment Share on other sites More sharing options...
mbayer Posted September 14, 2006 Report Share Posted September 14, 2006 I just finished installing and then re-installing Retrospect Workgroup 6.1.126 on a brand new Intel Mac Mini Core Duo 1.66 Ghz with a Fujitsu MHV2060BHPL SATA internal 60GB Drive running Mac OS X Server 10.4.7 (Build 8K1079). I did a full erase and install of Mac OS X Server, setup a few accounts, and then ran the Retrospect installer and rebooted. Launched Retrospect to configure it. After entering the license key I attempt to define a subvolume on the boot drive so I click on the "Configure" tab in the Directory window and then click the "Volumes" button. In the "Volumes Database" window I click on "Macintosh HD" and then click the "Subvolume…" button. At this point Retrospect just quits. Restarting Retrospect requires entering the the serial number again. Clicking the "Subvolume" button with the boot volume selected in ANY window where it is available results in the same behavior--Retrospect just quits. There is no explanatory entry in any log. If I attach a USB flash drive, I *can* select it and create a subvolume without any problem. This confirms that the problem is with Retrospect on OS X Server 10.4.7 on Intel and SATA drives. I too have a very anxious customer. Is there any kind or workaround? Thanks! Link to comment Share on other sites More sharing options...
Mayoff Posted September 14, 2006 Report Share Posted September 14, 2006 A bug has been opened with Apple and they are investigating. We are doing our best to come up with a workaround, but due to the nature of the problem a workaround will not be as good as an actual bug fix in the operating system. Link to comment Share on other sites More sharing options...
mbayer Posted September 14, 2006 Report Share Posted September 14, 2006 If you've been able to isolate a particular OS bug, are you now able to tell us what will and will not work reliably with the currently available builds? This would save me a lot of time that I'll otherwise spend experimenting. Thanks Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.