linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Machek <pavel@suse.cz>
To: Matthew Dobson <colpatch@us.ibm.com>
Cc: Paul Jackson <pj@sgi.com>,
	clameter@engr.sgi.com, linux-kernel@vger.kernel.org,
	sri@us.ibm.com, andrea@suse.de, linux-mm@kvack.org
Subject: Re: [patch 3/9] mempool - Make mempools NUMA aware
Date: Sat, 28 Jan 2006 09:16:42 +0100	[thread overview]
Message-ID: <20060128081641.GB1605@elf.ucw.cz> (raw)
In-Reply-To: <43DAC222.4060805@us.ibm.com>

Hi!

I'll probably regret getting into this discussion, but:

> > Or Alan's suggested revival
> > of the old code to drop non-critical network patches in duress.
> 
> Dropping non-critical packets is still in our plan, but I don't think that
> is a FULL solution.  As we mentioned before on that topic, you can't tell
> if a packet is critical until AFTER you receive it, by which point it has
> already had an skbuff (hopefully) allocated for it.  If your network
> traffic is coming in faster than you can receive, examine, and drop
> non-critical packets you're hosed.  

Why? You run out of atomic memory, start dropping the packets before
they even enter the kernel memory, and process backlog in the
meantime. Other hosts realize you are dropping packets and slow down,
or, if they are malicious, you just end up consistently dropping 70%
of packets. But that's okay.

> I still think some sort of reserve pool
> is necessary to give the networking stack a little breathing room when
> under both memory pressure and network load.

"Lets throw some memory there and hope it does some good?" Eek? What
about auditing/fixing the networking stack, instead?

> >  * this doesn't really solve the problem (network can still starve)
> 
> Only if the pool is not large enough.  One can argue that sizing the pool
> appropriately is impossible (theoretical incoming traffic over a GigE card
> or two for a minute or two is extremely large), but then I guess we
> shouldn't even try to fix the problem...?

And what problem are you trying to fix, anyway? Last time I asked I
got reply around some strange clustering solution that absolutely has
to survive two minutes. And no, your patches do not even solve that,
because sizing the pool is impossible. 
								Pavel
-- 
Thanks, Sharp!

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  parent reply	other threads:[~2006-01-28  8:16 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20060125161321.647368000@localhost.localdomain>
2006-01-25 19:39 ` [patch 1/9] mempool - Add page allocator Matthew Dobson
2006-01-25 19:39 ` [patch 2/9] mempool - Use common mempool " Matthew Dobson
2006-01-25 19:40 ` [patch 4/9] mempool - Update mempool page allocator user Matthew Dobson
2006-01-25 19:40 ` [patch 5/9] mempool - Update kmalloc mempool users Matthew Dobson
2006-01-25 19:40 ` [patch 6/9] mempool - Update kzalloc " Matthew Dobson
2006-01-26  7:30   ` Pekka Enberg
2006-01-26 22:03     ` Matthew Dobson
2006-01-25 19:40 ` [patch 8/9] slab - Add *_mempool slab variants Matthew Dobson
2006-01-26  7:41   ` Pekka Enberg
2006-01-26 22:40     ` Matthew Dobson
2006-01-27  7:09       ` Pekka J Enberg
2006-01-27  7:10       ` Pekka J Enberg
2006-01-25 19:40 ` [patch 9/9] slab - Implement single mempool backing for slab allocator Matthew Dobson
2006-01-26  8:11   ` Pekka Enberg
2006-01-26 22:48     ` Matthew Dobson
2006-01-27  7:22       ` Pekka J Enberg
2006-01-25 23:51 ` [patch 1/9] mempool - Add page allocator Matthew Dobson
2006-01-25 23:51 ` [patch 3/9] mempool - Make mempools NUMA aware Matthew Dobson
2006-01-26 17:54   ` Christoph Lameter
2006-01-26 22:57     ` Matthew Dobson
2006-01-26 23:15       ` Christoph Lameter
2006-01-26 23:24         ` Matthew Dobson
2006-01-26 23:29           ` Christoph Lameter
2006-01-27  0:15             ` Matthew Dobson
2006-01-27  0:21               ` Christoph Lameter
2006-01-27  0:34                 ` Matthew Dobson
2006-01-27  0:39                   ` Christoph Lameter
2006-01-27  0:44                     ` Matthew Dobson
2006-01-27  0:57                       ` Christoph Lameter
2006-01-27  1:07                         ` Andi Kleen
2006-01-27 10:51                   ` Paul Jackson
2006-01-28  1:00                     ` Matthew Dobson
2006-01-28  5:08                       ` Paul Jackson
2006-01-28  8:16                       ` Pavel Machek [this message]
2006-01-28 16:14                         ` Sridhar Samudrala
2006-01-28 16:41                           ` Pavel Machek
2006-01-28 16:53                             ` Sridhar Samudrala
2006-01-28 22:59                               ` Pavel Machek
2006-01-28 23:10                                 ` Let the flames begin... [was Re: [patch 3/9] mempool - Make mempools NUMA aware] Pavel Machek
2006-01-27  0:23   ` [patch 3/9] mempool - Make mempools NUMA aware Benjamin LaHaise
2006-01-27  0:35     ` Matthew Dobson
2006-01-27  3:23       ` Benjamin LaHaise
2006-01-28  1:08         ` Matthew Dobson
2006-01-25 23:51 ` [patch 7/9] mempool - Update other mempool users Matthew Dobson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20060128081641.GB1605@elf.ucw.cz \
    --to=pavel@suse.cz \
    --cc=andrea@suse.de \
    --cc=clameter@engr.sgi.com \
    --cc=colpatch@us.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=pj@sgi.com \
    --cc=sri@us.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox