logo       

Rebuild speeds and recently failed devices oops: msg#00071

linux.raid

Subject: Rebuild speeds and recently failed devices oops

Andrew,

I had set the min to 15000 (15MB/s) and max to 80000 (80MB/s). My
rebuild rate still wouldn't go above 7MB/s. :(

Sean.

> I was seeing similar rebuild performance on my large number of RAID1
> arrays (the first rebuilds went speedy and subsequent ones slowed down
> after the first ones completed). I bumped up the speed_limit_min and
> the rebuild performance went up to where I expected it to be.
>
> What were you speed_limit_[min|max]?
>
> -----Original Message-----
> From: Sean Kormilo [mailto:skormilo@xxxxxxxxxxxxxxxxxx]
> Sent: Thursday, March 13, 2003 8:48 AM
> To: Neil Brown
> Cc: linux-raid@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH] md - 1 of 12 - Missing mddev_put in md resync code
>
>
> Neil,
>
> >
> > No. There have been lots of changes since 2.4. Some patches might
> > conceptually apply, but they would need a bit of work to make them
> > actually apply.
>
> Ok, then I won't give it a try. Thanks for the heads up.
> >
> > >
> > > It looks like one or more of these might reduce the chances of the
> > > kernel oops I experience on occasion when manipulating arrays.
> >
> > Would this be when removing a recently-failed device? That can still
> > oops. I should send in a patch for that...
>
> Yes in fact, that is the one I was seeing.
>
>
> > These patches are about management, not performance. They won't
> > affect rebuild performance at all. It is already quite good in 2.4 I
> > believe. But maybe my experience is different from your.
>
> I have an odd problem where the first rebuild performs as expected
> (about 40MB/s), but any subsequent rebuild plods along at about 6-7MB/s.
> There is no IO activity on the system other than the rebuild during my
> tests.
>
> I tried putting in a couple of debug printk's here and there in the code
> that backs off the resync, and the printk's themselves returned the
> performance back to expected levels... some kind of a timing problem
> perhaps?
>
> Sean.
--

Sean C. Kormilo, STORM Software Architect, Nortel Networks
email: skormilo@xxxxxxxxxxxxxxxxxx


-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html



<Prev in Thread] Current Thread [Next in Thread>
Google Custom Search

News | FAQ | advertise