Subject: Re: Interactive responsiveness under heavy I/O load
To: None <tech-kern@netbsd.org>
From: John Goerzen <jgoerzen@complete.org>
List: tech-kern
Date: 01/26/2004 17:16:19
Thanks for your reply and explanation, Thor.  I have a couple of
follow-up comments:

On 2004-01-26, Thor Lancelot Simon <tls@rek.tjls.com> wrote:
> On Mon, Jan 26, 2004 at 04:30:08PM +0000, John Goerzen wrote:
> The new I/O sorting algorithm in -current should make this significantly
> better.  I am hoping that it can become the default for 2.0.

If I were to upgrade to current, how would I enable this sorting
algorithm on my system?

>> It seems like when this problem occurs, an I/O scheduler somewhere is
>> starving everything but the big writing process of resources.  But I
>> have no idea if this is tweakable somewhere, or how to go about fixing
>> it.
>
> You say later that you're using softdep.  The likely problem is an
> interaction of softdep and the questionable behaviour of the delayed
> write scheduling code (the "smooth-sync" or "syncer") code that was
> imported along with softdep.

[snip] 

> So there is a light at the end of the tunnel.  If you want immediate
> relief, turning off softdep should make your system's interactive
> performance more predictable, though it will probably make your
> I/O itself slower.

Significantly.  One of the first things I did with my new NetBSD system
was to untar pkgsrc.tar.gz.  It was S L O W.  I didn't make formal
benchmarks, but after turning on softdep, I'd say the performance was at
least four times better, if not more (maybe even twice that).

Have I stumbled across the reason softdep is not enabled by default, or
is there some other logic behind this?

While we're on the topic, does anyone have a nice comparison of ffs or
lfs to reiserfs somewhere?  I have found a number of (dated) comparisons
of ffs to ext2.

-- John