July 23, 2009 at 4:11 pm
well guys I am evaluating the product and ease of use and compression ratios are as promised, but the backup run times have doubled compared to the native backups, which is disappointing to say the least. I take it this does not fit in with your experiences?
Striping and using maxtransfersize have been suggested to bring run times back down to native backups but I do not want to use backup devices and it goes against the promise of not having to amend current backup processes.
---------------------------------------------------------------------
July 23, 2009 at 4:27 pm
george sibbald (7/23/2009)
well guys I am evaluating the product and ease of use and compression ratios are as promised, but the backup run times have doubled compared to the native backups, which is disappointing to say the least. I take it this does not fit in with your experiences?Striping and using maxtransfersize have been suggested to bring run times back down to native backups but I do not want to use backup devices and it goes against the promise of not having to amend current backup processes.
Something doesn't sound right. My backup times were reduced by 40 to 50%. One question I have, are you backing up to a local disk or a network share on another server?
Also, I'd consider contacting HyperBac, drop them an email. They are quite responsive, and you may find yourself talking to actually people on their dime. I had a minor issue when I first evaluated the product that had noting to do with the backups and they called me from Australia during my business hours here in the US.
July 23, 2009 at 4:32 pm
I do know that the data compression I get on our HR database is not as good as on the Finance database, but that I attribute to all the BLOB files uploaded as part of our TAM System we implemented a year ago. But still, our backups are 50% smaller than a native backup.
July 23, 2009 at 4:40 pm
backups were to local disk lynn.
I did email hyperbac and they responded promptly, striping and maxtransfersize suggestions were theirs, but that does not really cut it and should not be necessary, so interested if others hit this.
I an sending them the service log file, hope they find something.
I will try it on a different server,see if results are comparable.
I go on holiday saturday, didn't need this!
---------------------------------------------------------------------
July 24, 2009 at 8:42 am
george sibbald (7/23/2009)
backups were to local disk lynn.I did email hyperbac and they responded promptly, striping and maxtransfersize suggestions were theirs, but that does not really cut it and should not be necessary, so interested if others hit this.
Just curious what backup compression option you picked... the hbe fwiw encrypts as well as compresses... I'd expect backup times to increase with that. We just used the native zip compression and got the 40% reduction in backup times.
July 24, 2009 at 9:09 am
mtassin (7/24/2009)
george sibbald (7/23/2009)
backups were to local disk lynn.I did email hyperbac and they responded promptly, striping and maxtransfersize suggestions were theirs, but that does not really cut it and should not be necessary, so interested if others hit this.
Just curious what backup compression option you picked... the hbe fwiw encrypts as well as compresses... I'd expect backup times to increase with that. We just used the native zip compression and got the 40% reduction in backup times.
no encryption, basic compress to zip format. tried 3 servers, 20 databases, run times increased by 50%. This is version 4.1.0.0 (latest version). I added .bak as a file extension and set parameters for that.
the only thing that brings run times back down is striping, but I should not have to do that and its not an option in a maintenance plan for multiple databases.
---------------------------------------------------------------------
July 24, 2009 at 9:26 am
Is CPU pegged while the backups are running?
Do the databases contain a lot of information that is already compressed or unsuitable for compression?
What compression ratio do you get if you take a standard backup and zip it?
July 24, 2009 at 9:46 am
This is really weird. Is there anything else running on the servers? Anti-Virus software, etc?
July 24, 2009 at 9:50 am
matt stockham (7/24/2009)
Is CPU pegged while the backups are running?
total processor % jumps up about 25% to 50% average, speread evenely across the cpus (4)
Do the databases contain a lot of information that is already compressed or unsuitable for compression?
if you mean text, varbinary, no, standard stuff
What compression ratio do you get if you take a standard backup and zip it?
slightly better, but very much on a par with hyperbac. (80%)
I'm beginning to feel persecuted, I had my hopes pinned on this!
---------------------------------------------------------------------
July 24, 2009 at 9:58 am
george sibbald (7/24/2009)
matt stockham (7/24/2009)
Is CPU pegged while the backups are running?total processor % jumps up about 25% to 50% average, speread evenely across the cpus (4)
Do the databases contain a lot of information that is already compressed or unsuitable for compression?
if you mean text, varbinary, no, standard stuff
What compression ratio do you get if you take a standard backup and zip it?
slightly better, but very much on a par with hyperbac. (80%)
I'm beginning to feel persecuted, I had my hopes pinned on this!
Getting weirder. My systems don't spike that much when doing backups. We hardly notice the difference when ours are running. The directory that you are backup up to isn't compressed is it?
July 24, 2009 at 10:00 am
Lynn Pettis (7/24/2009)
This is really weird. Is there anything else running on the servers? Anti-Virus software, etc?
McAfee is on all the servers, should be configured to avoid SQL files though, and doesn't bother native backups. :unsure:
---------------------------------------------------------------------
July 24, 2009 at 10:05 am
I've tried on 3 different servers, can't all be bad. common denominator is hyperbac. what version you guys on?
---------------------------------------------------------------------
July 24, 2009 at 10:20 am
george sibbald (7/24/2009)
I've tried on 3 different servers, can't all be bad. common denominator is hyperbac. what version you guys on?
I'm still running 3.0.11.0.
July 24, 2009 at 10:20 am
3.0.22.0
We probably should upgrade it... but in the realm of "when it works" we tend to leave things like this alone.
July 24, 2009 at 10:30 am
....then perhaps we have a bug................in a way I hope so
---------------------------------------------------------------------
Viewing 15 posts - 16 through 30 (of 55 total)
You must be logged in to reply to this topic. Login to reply