on Jun 22, 2015 at 11:15:59 pm Last Edited By Trevor Cable on Jun 22, 2015 at 11:34:54 pm
This morning our Terrablock 24D went down and we lost a volume. Support got us back up in about 45 minutes but our Director had to leave his edit session because of this. We were unable to recover our lost volume (the volume was still there but was now empty) but keep regular backups so after a quick unarchival we were back up and running. Our Director (large animated motion picture) was not happy and I certainly heard about it.
So as we have two Terrablocks with plenty of storage on the second I am interested in finding a good way to mirror volumes across the two. Not in real time probably but I'm thinking every 4 hours an rsync would happen. At the moment for my backups I have a client machine with a dual 10Gb card connected to each TB and a large RAID. This is kept in another building from the Terrablocks. I use a shell script to mount and copy each volume using rsync.
As I have my Terrablocks each connected to the same switch is there a way to do an incremental backup from one system to another? Can the Terrablocks see and access each other? I'll run this by support but thought maybe someone here has done something similar.
I heard about your crash yesterday, it's an error we haven't seen before and we'll have more information once we look through the logs. As for the volume - in the rare case where a file system is corrupted, it can often be quicker to run a restore application than restore from the archive. The files are all still there, the pointers are wrong.
For the future, yes the TerraBlock servers can access each other, through any method you like - FC, 10Gb, etc. We can get you a license for the synch software we use on the servers that can do exactly what you're looking for.
Thanks for the feedback. I believe had I run a disk check on my corrupt volume I would not have lost my media. I did this before support had a chance to recommend I run a disk check. As it turns out the index was corrupt and was most likely fixable. My bad for being impatient.
I appreciate you guys checking the logs see to what happened. As always your support team is fantastic.