Jump to content

No response, kernel panic, reboot server


Recommended Posts

Our department just bought a brand new PowerMac G5 1.8GhzMPP with Mac OS X Server 10.3. We also purchased a new VXA-2 FireWire drive from Exabyte. I installed Retrospect 6 on the new machine (thanks to the ASM) expecting no problems like I had running Retrospect 5 on an older machine with a SCSI DAT drive.

 

Boy, was I wrong.

 

I've added clients to the backup - both Windows and Mac. I have createda a backup script and let it run overnight on more than one occasion. The results have been the same: Retrospect appears to be finishing up a job and claims that it is done at the time I specify. However, the program, does not respond. Even if I wait 30min after it should have been done with the backup script, it's still unresponsive. So I force-quit the program and try and relaunch it. When I do a device scan, it causes a kernel panic. So I have to reboot the server and explain to my users why the new server has gone down twice in as many days. I also need to explain why this new backup system isn't running yet, despite spending over $5,000 to implement it.

 

I would appreciate some answers.

Link to comment
Share on other sites

Quote:

I have to reboot the server and explain to my users why the new server has gone down twice in as many days. I also need to explain why this new backup system isn't running yet, despite spending over $5,000 to implement it.

 


 

 

 

To be brutally honest, I'm surprised you haven't also been asked to explain why you deployed a new server before fully qualifying it in pre-production testing. Certainly these glitches should have been discovered _before_ moving users' files.

 

 

 

- What happens if you _don't_ force quit the program? Perhaps it needed more then the 30 minutes you allowed for it to exit gracefully.

 

 

 

- What's the memory situation with the new machine? Macintouch did a whole special report a few weeks back regarding the effects of bad RAM on OS X. It may be too early to suspect this, but it should always be considered a possible culprit.

 

 

 

- There have been some comments on the Dantz RetroTALK mailing list about VXA-2 drives and Retrospect 6, but no reports of OS crashes. You might try 5.1 on the same machine and see if it's better behaved.

 

 

 

- Did you try unplugging the FireWire cable after you quit the program and then replacing it before restarting Retrospect and attempting to have it scan for devices again? If the kernal panics are not memory related then they're happening at a very low software level; allowing the OS to get a fresh look at the device before asking Retrospect to control it might be helpful.

 

 

 

- You should open an incident with Dantz technical support. If it turns out that there is a bug in 6.0 that is causing it to fail with VXA-2 drives, they'll refund your support fee (on request). They're very fair in this regard (and it's not all that much money up front).

 

 

 

Dave

Link to comment
Share on other sites

CallMeDave said:

 

Quote:

To be brutally honest, I'm surprised you haven't also been asked to explain why you deployed a new server before fully qualifying it in pre-production testing. Certainly these glitches should have been discovered _before_ moving users' files.

 


 

 

 

Granted, I should have setup the new machine for backups only. OTOH, Dantz did say that the hardware and software combination I was implementing should work without any problems. I suppose it's my fault for assuming Q.A. on a new release of software.

 

 

 

Quote:

- What happens if you _don't_ force quit the program? Perhaps it needed more then the 30 minutes you allowed for it to exit gracefully.

 


 

 

 

Well, considering it took over 48 hours to rebuild a catalog file from a 15GB tape (and it STILL wasn't done), I'd say that the problem isn't in giving Retrospect 6 enough time.

 

 

 

Quote:

- There have been some comments on the Dantz RetroTALK mailing list about VXA-2 drives and Retrospect 6, but no reports of OS crashes. You might try 5.1 on the same machine and see if it's better behaved.

 


 

 

 

That's what I'm doing for now. I only hope that the new hardware - new Mac and new tape drive - will help me avoid the catalog corruption issues that led me to install Retrospect 6 in the first place. I really can't stop backups every few days in order to rebuild corrupt catalog files.

 

 

 

Quote:

- You should open an incident with Dantz technical support. If it turns out that there is a bug in 6.0 that is causing it to fail with VXA-2 drives, they'll refund your support fee (on request). They're very fair in this regard (and it's not all that much money up front).

 


 

 

 

We'll see how far I get on here and on the mailing list.

Link to comment
Share on other sites

  • 3 weeks later...

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...