No more hiding in the server room
Backup systems, HW & SW
Most the places I visit use symantec backup exec and tape drives.
I hate tapedrives, detest them, I've replaced nearly 10 of the things in the last year.
I'm not overly fond of backup exec either, it seems to be to be overly complex, and when things go wrong with a backup it's not always easy to tell what the problem is.
other sites use RDX cartages, which I like better than tapedrives, but these can run into problems with the backup software not releasing the cartridge and not being able to eject it, (RDX is basically a USB drive) also symantec does not make it clear that before a backup with a removable usb drive, you need to do an inventory job. you can also still run into problems with the catalogue though.
we've got one client that uses CA ARCserv backup, which I detest even more than backupexec
that may be because I've not had as much to do with it, but it seems even MORE complex than backup exec is. also when trying to restore a specific file or what have you, it takes a good 15min to drill down to select the file you're wanting to recover.
I've yet to see a good solution that is low maintenance.
Views: 208
Replies to This Discussion
-
Permalink Reply by Scott Seltzer on -
I've worked with quite a few options, and Backup Exec seems like the least headache for me so far. We're tending to move away from tape, in favor of D2D backups.
The nice thing about BE, which I didn't really appreciate until I used other solutions, is how informative it can be. If you learn to use the tools that come with it, it will tell you an awful lot about why a backup isn't working. Some of the other systems I've used are more or less black boxes for troubleshooting purposes, and won't give any clues as to what's going on inside.
Honestly, my favorite backup system to maintain (I haven't used it in 5 or 6 years), was AMANDA. Set it up right, and it just worked. Aside from swapping tapes and test restores, I'd touch the system once or twice a year to make sure that backup sizes weren't outgrowing the disks.
-
Permalink Reply by Noj on -
BE is sometimes informative yes, but it can be un-necessarily cryptic. in other places it's not informative at all though:
for example a recent B2D failure, check the job history log and it doesn't give much more than
"Completed status: Canceled, timed out
The job was automatically canceled because it
exceeded the job's maximum configured
run time."
checked the disk it was backing up too, still 100gb free space on that so it should be fine, after a bit of scratching my head I checked the B2D folder properties and saw that it was set to fail if only 100gb left free on the disk. BE would be a lot better if it could tell me basic stuff like this in the job log.
BE is great when it's set up right and working. but it can be a pain to figure when it's not working.
are you using removable/usb disks for your B2D backups? have you ever come across problems restoring from them?
-
Permalink Reply by Scott Seltzer on
-
There should have been an alert generated when it hit that limit.
We're using Buffalo TerraStations in most cases, with USB drives backing up from them for offsite. I haven't seen any problems with restore jobs.
-
Permalink Reply by Noj on
-
Ah! an alert eh? now that I've gone looking for it, yes I see it. should probably set that up so it sends an email notification.
How do you set your jobs up for the USB drives?
Got one client who uses USB drives which get swapped out each week, first backup of the week is full, the rest incremental.
I need to sit down with that one and check it out, backups appear to be working fine, but when I try to do a test restore, I run into problems. possibly I need to do an inventory then a catalog on each drive... maybe delete the catalog and recreate it. time consuming.
two other sites use RDX, different cartridge for each day of the week. they seem to be going fine now we've set up an automatic inventory job before the backup job. took us a while to figure that one out.
-
Permalink Reply by Graycat on -
At the moment we're in a state of flux whilst we implement SANs on two sites and move some of our key servers over to a virtual platform but historically we've been pretty happy with BE and tapes.
Going forward we're planning on looking at things like VEEM for the virtual infra but some things are just easier with BE (ie: Enterprise Vault) so we'll never really get rid of it.I suppose the choice of media boils down not to how you do the backups but how you want to do the restores and what you want to be able to restore.
We use a lot of server-to-server replication for things like main data servers which is great if you stage things so they're slightly out of synch but if you need a file restoring back to last week / a fortnight ago / a month or even longer then you're going to struggle IMO. Our overall strategy is to replicate to another server locally (cheap HP NAS / storage server) and then a DR server on another site. The data is replicated to be 24hrs out of date down the line however we do full backups of the live server weekly with daily diffs to tape too. As a longer term restore point we take one set of tapes out every six months to provide a restore point. These are then kept in long term storage / archive off site.
The trick I've found with tapes is to make sure you get a good autoloader and keep it working sweetly coz when things start going awry it'll be a pain to sort! :) lol
-
Permalink Reply by Noj on
-
none of the sites I visit are large enough to worry about having an auto-loader, to me that sounds like adding yet another point of failure.
I'm now having a problem with a system that seems to have stopped compression, a week ago they had about 36gb left on the tape after a backup, now they're full. media tab in BE shows it's gone from 1.1:1 compression to 1.1 a quick mental add-up of the bytes backed up a week ago and now only shows a few gb more being backed up. was set to HW compression, ran HP tape tools, which reported that it was compressing ok. tried it on SW compression with no change.
removed a bunch of un-necessary stuff from being backed up, will see if I've removed enough.
not making a lot of sense though.
© 2013 Created by Elizabeth Ayer and Michael Francis.
Powered by
