Les Mikesell <email@example.com> wrote on 09/17/2012
> On Mon, Sep 17, 2012 at 11:54 AM, Timothy J Massey
> <firstname.lastname@example.org> wrote:
> > However, I have recently inherited a server that is >3TB big,
> > full, too! Backups of that system take 3.5 *days* to complete.
> > live with that. I need better performance.
<Quick fixes snipped>
> In any case you have to keep in mind that if
> have to restore, there will be considerable downtime. For the
> of a disk these days, it might be worth keeping a copy up to date
> native rsync that you could just swap into place if you needed it
> back that up with backuppc if you want to keep a history.
Fortunately, BackupPC is a "backup of the backup"
right now, and is not expected to be used for real. Yet. That's
why I can take the time and try to actually solve the problem, rather than
But that will likely end in November, if not sooner.
> > It looks like everything is under-utilized.
For example, I'm getting a
> > measly 40-50MB of read performance from my array of four drives,
> If every read has to seek many tracks (a likely scenario), that's
> unreasonable performance.
Which is why I asked if others were getting better
performance. They are. So it's not just inherent to the task.
There's something different.
> My physical drive and network
> > lights echo this: they are *not* busy. My interrupts
> > manageable and context switches are very low. Even my CPU
> > tremendous: nearly no time in wait, and about 50% CPU idle!
> I think you should be seeing wait time. Unless perhaps you have
> huge files that end up contiguous on the disk, I'd expect the CPU
> be able to decompress and checksum as fast as the disk can deliver
> and there shouldn't be much other computation involved.
I'm not. I've got two of these new boxes built.
In both cases, they have 2-4% wait time when doing a backup. One
is a RAID-5 and one is a RAID-6.
Might it have something to do with md? Could
the time that would normally be considered wait time for BackupPC be counted
as CPU time for md? That doesn't seem logical to me, but I can say
that there just isn't any wait time on these systems.
> Mine seem to track the target host disk speed more than anything else.
> The best I see is 208GB with a full time of 148 minutes.
> is with backuppc running as a VM on the East coast backing up a target
> in California and no particular tuning for efficiency. Compression
> on and no checksum caching.
That's the same settings I'm using. But that's
about double the performance I'm getting. 247GB in 340 minutes, or
I've just turned compression off for a couple of hosts.
We'll see how this affects their performance. I'll let everyone