Jump to content

Retrospect hangs Exabyte VXA-2 drive


Recommended Posts

G5 Xserv running OSX 10.4.6

Exabyte VXA-2 packet loader tape library

 

Awhile back I started to receive 102 errors during backups to my exabyte VXA-2 tape library. After researching the issue I upgraded the firmware to v. 210D for the drive and V1A10A for the autoloader. I also upragded from Retrospect 6.0 to Retrospect 6.1. That seems to have fixed the issue for about 4 backup runs. Then I started to receive 205 errors , " lost access to storage medium", early into the write part of the backup. When this happens Retrospect hangs and I have to force quit the process and restart Retrospect. It seems that any attempt to use the tape device results in a 205 error and or hung Retrospect application. Any ideas?

Link to comment
Share on other sites

I suspect that I've got a similar, if not amost identical, configuration as you do, and I'm not having any problems. A few bits of important information that you have not provided:

 

(1) is this the SCSI or Firewire version of the Exabyte autoloader? I've got SCSI, can't help if you've got Firewire.

 

(2) if SCSI, what SCSI interface card are you using? I've got an ATTO UL4D; had problems with the Apple BTO SCSI card (really, an LSI Logic 22320) that disappeared when replaced with ATTO.

 

(3) if SCSI, what other devices are on this SCSI channel?

 

(4) if SCSI, I assume/hope that you've got good LVD cables and terminator.

 

(5) EXACTLY what Retrospect RDU and Retrospect version do you have (see Retrospect log file) ?

 

(6) what does Retrospect's "Configure > Devices" show when in the failure mode?

 

(7) what does Retrospect's "Configure > Devices > Device Status" show when in the failure mode?

 

(8) what does Apple System Profiler on the server show for the tape drive and autoloader?

 

(9) have you tried cleaning the heads with a cleaning tape and using a new tape?

 

(10) Are you sure you didn't make a typo on your firmware upgrade version for the autoloader? ("V1A10A for the autoloader") - the current autoloader firmware is "V1A10B" (09/26/2005) and the prior version was "V1A10A" of 08/29/2005. However, it appears that the 09/26/2005 V1A10B update only matters when the autoloader is used with the newer VXA-320 drive (overtemp sensing).

 

(11) Have you tried to run Exabyte's diagnostics from their, um, lovely vxatool ? Be sure to use a scratch tape because the diagnostic will put down a tape header that makes Retrospect unhappy, and will destroy data on the tape.

 

(12) Regarding the "hangs" - are you able to ssh in (or ARD in) and look at running processes, etc? What does that show? Might indicate whether there is a block on pending I/O or whether there is a user-level issue.

 

(13) Regarding the "hangs" - Has Retrospect yet moved away from BOT with writes, or is this the first write to the tape when Retrospect is doing its device scan to discover devices?

 

Simply FYI, it was for situations like this that I got the Ethernet admin interface for the autoloader. Don't use it except in situations like these, but it gives independent information as to what is happening with the autoloader/drive when I can't figure it out from the server side, and has been helpful. Interesting too that it emails me when something happens (such as someone manually inserting/removing tapes, etc.).

 

Provide some more info and I will see if I can make some suggestions. I saw a hang a while back and was able to devise a workaround which might be applicable to your situation if I knew more details.

 

Regards,

 

Russ

Link to comment
Share on other sites

Thanks for the information. I am using the Firewire interface for the VXA-2 library.

 

(5) The the Retrospect version lists as 6.1.126. I've noted that since the upgrad the RDU no longer lists with the Retrospect version. Prior to the Retrospect upgrade from 6.0.212 to 6.1.126 the RDU version was 6.3.102.

 

(6) & (7) when in failure mode, the Retrospect application is hung.

 

(8) System Profiler report folows

VXA-2a:

 

Manufacturer: Exabyte

Model: 0x0

GUID: 0x10100319002096

Maximum Speed: Up to 800 Mb/sec

Connection Speed: Up to 800 Mb/sec

Sub-units:

VXA-2a Unit:

Unit Software Version: 0x10483

Unit Spec ID: 0x609E

Firmware Revision: 0x201

Product Revision Level: 210D

Sub-units:

VXA-2a SBP-LUN:

VXA 1x10 1U Unit:

Unit Software Version: 0x10483

Unit Spec ID: 0x609E

Firmware Revision: 0x201

Product Revision Level: A10B

Sub-units:

VXA 1x10 1U SBP-LUN:

 

(9). I've used a new cleaning tape and data tape with the same failure resulting

 

(10) that was a typo the firmware level is A10B

 

(11) VXAtool reports no error. the firmare update reported no errors.

 

(12) I'm connected via ARD and the Activity monitor does not indicate any bottleneck

 

(13) The tape starts to write, according to the Retrospect monitor window. After a few files are listed then the 205 error occurs.

 

Where can I get the Ethernet admin interface?

 

Thanks for your help

Link to comment
Share on other sites

Quote:

I am using the Firewire interface for the VXA-2 library.

 


Sigh, well, then my experience isn't applicable here on the Exabyte issues. Sorry.

Quote:

I've noted that since the upgrad the RDU no longer lists with the Retrospect version.

 


Really sounds like you need to install RDU 6.1.5.102 (4/17/2006) and that whatever RDU you have is not being loaded because it's too old for the Retrospect update you have. There's a link to the RDU here: EMC Retrospect Mac Updates

Quote:

(13) The tape starts to write, according to the Retrospect monitor window. After a few files are listed then the 205 error occurs.

 


Ok, you've got something different than what I saw with the SCSI version with other SCSI devices (RAID 1 mirrors controlled by SoftRAID) on the same channel as Retrospect's tape drive. My experience is not helpful here because you've got Firewire.

Quote:

Where can I get the Ethernet admin interface?

 


I don't think that it is available with the Firewire version, for some odd reason. It was an optional add-on ("PacketLoader Plus") for the VXA-2 SCSI version, and it's now standard with the VXA-320 SCSI version. The only time I've used it is during Retrospect troubleshooting, when the SCSI chain was hung and I was trying to figure out what was happening. But I do urge that you get the RDU update. Some of the EMC/Dantz experts might be able to help better than I can. Dantz has the Firewire version of this drive/autoloader in their labs for testing, so they are probably better help than I can be. If the RDU update doesn't help things, I suggest you pay for a support call or purchase the support contract - we've found their support group to be worth it, and you don't get stuck with the usual front-line people at other vendor tech support groups who ask if the computer is turned on. However, many of the support people visit these forums, so perhaps they have seen your problem before. Other than the RDU update, sounds like you have tried all the right things. I suggest you try the RDU update first. Regards, sorry I couldn't be more help. Russ

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...