Graeme Humphries wrote:
On Wed, 2005-06-29 at 23:45 +0200, Stefan G. Weichinger wrote:
Yeah, and they are *really* static. There's very little that changes
day-to-day compared to the overall backup size. I'd estimate we have
less than 1% change on the larger shares. This should work awesome, time
to run a test backup. :)
Let us know, maybe this motivates more people to move on to 2.4.5.
Looking at my results from yesterdays backup with "estimate server", and
it's shaved *at least* 5 hours off the run! I am a very happy man. :D
That's the way it should be ;)
Now, looking at the graphs that amplot generates (which is going to be
invaluable to us, kudos for making that utility!),
I'd like to see some improvements on that tool, but in general it's very
useful, yes.
I think I can
probably shave another 2 hours off now that it's got one full run of
stats and can better utilize the "HTHTs" scheduling policy I put in
place. :)
Make sure to let things settle in first before attempting too much
optimization to fast. AMANDA adapts to your setup by "learning from
history", commonly spoken. Give it some runs of amdump to see the trends ...
Of course, now I'm running into a problem where the crappy Silicon Image
SATA controller I've got my scratch disks hooked up to dies horribly
under heavy load, but that's hardly AMANDA's fault. ;)
You're right about that ;)
Greets, Stefan
--
Stefan G. Weichinger
AMANDA core team member
mailto://[EMAIL PROTECTED]
--
oops! linux consulting & implementation
http://www.oops.co.at
--
