We are testing CatDV Workgroup Server and believe it could be a nice addition to our workflow.
One important area that is creating challenges for us is archiving footage to multiple SATA drives. Yes, a Cache-A LTO appliance is more than likely a better solution, but we are recovering from an awful 5 year experience with a Quantum LTO3 3A Superloader. For me to suggest spending 10K for another LTO appliance at this point in time is not feasable.
We are using a CRU-DataPort RAX 210-QJP storage enclosure to archive to. It is a rackmountable SATA Drive backup unit whith two removable SATA Drive ports. It allows you to easily eject and insert new SATA drives without hassles.
For our testing I am using two SATA drives. I named them VAULT1000A & VAULT1001A
I archived footage on both drives and then purged the media from our SAN.
I then selected 4 video clips to restore. 3 clips on VAULT1001A and 1 Clip n VAULT1000A
When I “Restore Media Files” it will always ask for the Hard Drive that is not mounted.
I then eject the drive it is not asking for and mount the drive it is asking for.
Note: I am not mounting both drives simultaneously to see how CatDV deals with a drive that is not mounted
When I do this, it does not restore the files, but instead asks for the other drive that is not mounted.
It appears CatDV wants all drives mounted simultaneously before it will restore the media.
This is not a big deal if I only have a couple drives, but if I have several archive drives then this is not going to work unless I find a workaround.
Does this work correctly with one drive? One thing to note is that the Archive plugin only stores one set of archive data so you might need to clone the drives after they are done and just store one set of archive metadata.
Also note that you could consider just naming the drives the same thing at the OS level and labeling them differently on the outside, that way the system should restore no matter what. This might solve your problem with the current workflow. Paths need to be maintained or things get very wonky.
I think this could be accomplished far easier and more reliably with a set of Worker Node actions.
Each transfer to the volume could be logged in a separate field, for instance "archive transfer 1" and "archive transfer 2" could have separate metadata fields. A worker action for each archive could be set up, and they'd report back asset by asset. Restores would be far easier as well as you could record the path on each drive and then the system could pull as needed.
You could even add a checksum and verify to run afterhours if you were using the Worker.
It should prompt you about any drives that aren't loaded but will continue and restore the other available files if you press Ok. Then you can mount the other drive (and choose it as the archive location if necessary) and restore from that one.