Link to home
Start Free TrialLog in
Avatar of KalelGC
KalelGCFlag for United States of America

asked on

Backups running slow

Currently we are backing up about 700gb of data on the weekends and our backups are lasting almost 36 hours. We were previously running and SDLT unit in conjunction with an LTO3 which was our main work horse. In an attempt to speed things up, we purchased a 2 drive LTO 3 unit (with only one of those drives currently in production). We appear to be getting very similar performance on the new unit, which makes me believe we have a bottleneck somewhere else. What doesnt make sense to me is our servers are all running through the same switch, but some of the more souped up servers back up slower than the older servers. SQL backs up super fast. Any suggestions as to what I can look into next.
Avatar of L3370
L3370
Flag of United States of America image

Got a few questions here.

First, does your organization operate 24/7?  Would there be any reason to have a high volume of network traffic durring the backup process?

What have your backup job logs shown?  Backups can slow down when trying to copy files that are currently open or in use.  If advanced open file option is purchased, it will open and resume copy. If not it will skip the file.  Which brings me to another question...Does it seem that there are a lot of skipped files?

Also, are their any files that failed to backup?  I know a problem I have on my network is users saving PST's on the network even though I tell them not to.  PST get huge and corrupt, and end up making my backups take longer than desired.

If your organization is NOT 24/7, you could browse to your server's file shares and close all opened files on the network. That may help. Encouraging/mandating your users to log off when leaving could help remedy open files as well.

Sometimes I also get some fails because Backup exec could not get the desired "quiet time" on the resource to perform the backup.  Reducing server load could help...if it is possible.

I have a network that backs up roughly the same amount of data as you have described. There have been many times that it took 18-20+ hours.  I hate linear media, but thats the kind of garbage I have to put up with...
Avatar of KalelGC

ASKER

Correction to my original post. I was previously running and LTO2 and we purchased an LTO3 and are getting the same throughput.

L33370 - thank you for the response. Our company only runs 9-5. I have pinpointed some of the open/changing files that slow down the backups, but for example our exchange server only gets a throughput of 30mb/min, our server backups average to about 300mb/min. Then our SQL comes through at 2000mb/min. The 30mb and 300mb seem way to slow for me.

We are using AOFO. In regards to PST's, the server that hosts our PST files is one of our fastest servers to backup (1,700mb/min) It seems to me like some of our calmer servers give me the most problems.

Anymore questions or suggestions?
ASKER CERTIFIED SOLUTION
Avatar of L3370
L3370
Flag of United States of America image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of KalelGC

ASKER

Exchange is not using AOFO. In addition to the IS and SS, we also back up exchange mailboxes and public folders in order to allow quick one email restore when needed.  I understand that excahnge backups are slower than others. I guess what I need confirmation on, is do our backup speeds seem lower than what other people typically experience. The new backup unit we purchased is running fiber and we currently have a gig ethernet backbone and are getting nowhere near what i would expect on throughput.

We dont even have that much data we are backing up - 20GB on exchange I think?

Currently we are using unitelligent dell switches but will be upgrading to layer 3 cisco switches within the next few weeks. Do you think that will help things out, or will we still be bottlenecked at the servers themselves?
Your setup should be fine. I'd suggest upping the point value on this to 500. This will entice some of the Exchange Guru's on EE to take a look and suggest more options.

In the meantime...If you think it is a bandwidth/bottlenecking issue, try running a backup job on exchange and NOTHING ELSE. If the speed is still the same when there is no traffic from the other backups then it may not be a bottleneck issue.
Avatar of KalelGC

ASKER

problem solved