Continual "Can't access volume (error -1101)" on two Macs (10.6 & 10.7)

[theboyk]

I have two Macs in my 40+ pool of Macs that failed during a nightly backup script with a "Can't access volume" error (error code -1101). I'm wondering if anyone has any ideas why this keeps happening?

 

Server Details

- Intel Mac Pro running OS X Snow Leopard Server (10.6.8) & Retrospect Server 9.0.2 (unlimited)

- dedicated to running Retrospect Server

- dedicated bonded gigabit ethernet connection

- backing up to an external LTO-4 tape drive

 

Backup Details

- regular backup

- runs nightly at 10:30 PM

 

Client Details

- Intel iMac running OS X 10.6.8 & Retrospect 6.2.234

- Mac mini (mid-2011) running OS X 10.7.4 & Retrospect 9.0.2.102

- both machines are set to restart/wake 15 minutes before the backup runs (and are not set to sleep at all)

- both machines are on dedicated gigabit ethernet connections

- both machines have static IP addresses

- both machines have a single source folder that gets backed-up

- no firewall or anti-virus software running on these machines

 

I have tried the following to correct the issue:

- removed source folder from Retrospect Server (and re-added)

- removed source client from Retrospect Server (and re-added)

- uninstalled Retrospect client from machines (and re-installed)

- corrected permissions on each of the machines

- had the machines sitting on login window as well as logged in as a user prior to backup

 

No matter what I've tried, I can't get these two machines to consistently backup with success. The first time a backup runs after I've tried one of the above fixes, it'll backup with success, but then the following night, we're right back where we started with an error -1101 and a failed backup.

 

The only information in the (eye-hurting) logs is as follows:

 

Can't access volume XXX on XXX, error -1101 ( file/directory not found)

 

And in Retrospect, in the Sources list, the source folder that's been set as a favourite folder has what I can only describe as a small lightning-bolt icon to the left (which of course isn't mentioned anywhere in the Retrospect 8 manual). Yet, if I try an browse this folder (via Retrospect), I have no issue (and then the lightning-bolt icon goes away).

 

Does anyone have a clue why this continues to happen?

 

Thanks,

Kristin.

[Lennart_T]

Do you have any other (Mac) clients that has a source/favourite folder that does get backed up OK?

 

The main reason for getting the "Can't access volume..." error is adding the volume to the script instead of the client. Then if the user renames the volume, you get that error message.

[theboyk]

Yes, all of the 40+ machines are set to backup from a single source/favourite folder (per machine). In fact, 90% of the machines are built from the same custom installation image (for remote deployment/imaging) and contain more-or-less the same software (there are some minor differences in the version of the Adobe Creative Suite running on the machines, for example, but nothing fundamental that should cause an issue like this).

 

As well, the user doesn't have the ability to rename the volume itself nor the source/favourite folder itself.

[Daniels]

The other possibility is that the computer is asleep when the backup starts. Also make sure that the hard drive is not put to sleep and make sure a user is logged into the computer and the computer is not at the log-in screen.

[theboyk]

The other possibility is that the computer is asleep when the backup starts. Also make sure that the hard drive is not put to sleep and make sure a user is logged into the computer and the computer is not at the log-in screen.

 

Nope. As per my OP:

 

- both machines are set to restart/wake 15 minutes before the backup runs (and are not set to sleep at all)

- had the machines sitting on login window as well as logged in as a user prior to backup

 

I've seen it mentioned a couple times now that the user should be logged in. This is crazy. Retrospect 9 requires the computer to be logged in? This is a major security issue. It's out corporate policy that users log out at the end of each day — in fact, machines are scripted to restart and remain on the login window at a specific time (about 15 minutes prior to the backup process runs). Of the other 40+ Macs on the network, they all remain on the login screen during Retrospect backups and we have no issues with backups failing. That said, as above, I've had the user (as well as the admin user) logging in prior to backups, yet they continue to fail.