Our experience has been pretty close to Greg's. We're on our second Facilis server and encountered a lot of the failing drive issues that Greg mentioned.
I might go so far as to advise that you plan about three to four years out that you will likely have to swap out the drives for new ones. We discovered that not too long after the three year warranty, our drives starting dropping one by one. While this is mostly bearable fixing on a drive to drive basis, it did occasionally cause a problem on a busy day and, depending on how bad the recovery, it isn't always advised to do large writes to the server during a recovery. We discovered a bit by accident while a drive was rebuilding that it was creating corrupt frames in a few days worth of footage that was being put up. If I had it to do over again, I would have told my company to plan on buying a set of drives and taken a weekend to swap them out around the four year mark. It would have saved us a lot of headaches. Not Facilis fault necessarily but the cost of maintaining a server that large.
These recovery times have also been alleviated a bit by the latest server software which recovers MUCH quicker than it did before. We were on an old XP Pro server so that had it's limits.
We also ran into the dreaded two drives dead at once and Facilis spent A LOT of time dialed in and working on the server to get us to the most recent snapshot they could. This was able to stave off death for a while to allow us to archive some stuff. Eventually age overtook and the drives just couldn't keep up and we wound up losing some media. It wasn't a catastrophe but it could have been.
One workflow thing that we found curious: on our older XP server, multiwrite was basically unusable. Facilis proposed a workflow where a read only user could connect to the write user over the network (GigE) and drop files on their volume that way.
(If that's not clear, EDIT has write access. GFX does not. GFX connects to EDIT over the network and mounts EDIT'S Facilis volume as a network drive GFX then adds whatever files EDIT needs over the LAN and not over fibre.)
This was a great workflow for us. When we got the new server, we found out the hard way that using that workflow was not recommended and would in fact cause unmounting issues with any volumes that were written to that way. So we've had to change our workflow a bit.
We haven't jumped into the multiwrite volumes too much. They weren't fast enough for us on the older server and we have yet to really hammer one in our current workflow. By most reports, it is much more usable than it was when first introduced.
Hope that helps. Good luck.
Scott
__._,_.___
| Reply via web post | Reply to sender | Reply to group | Start a New Topic | Messages in this topic (25) |
.
__,_._,___
No comments:
Post a Comment