From: Pete Zaitcev <zaitcev@redhat.com>
To: Tony.P.Lee@nokia.com
Cc: linux-kernel@vger.kernel.org, zaitcev@redhat.com
Subject: Re: InfiniBand BOF @ LSM - topics of interest
Date: Tue, 14 May 2002 16:42:45 -0400 [thread overview]
Message-ID: <20020514164245.C13781@devserv.devel.redhat.com> (raw)
In-Reply-To: <4D7B558499107545BB45044C63822DDE3A206D@mvebe001.NOE.Nokia.com>
> Date: Tue, 14 May 2002 13:19:13 -0700
> From: <Tony.P.Lee@nokia.com>
Hrm. Didn't a guy from Nokia cooked a bizzare SM or an IB switch
software, which required ultra-fast mprotect?
> I like to see user application such as VNC, SAMBA build directly
> on top of IB API. I have couple of IB cards that can
> send 10k 32KBytes message (320MB of data) every ~1 second over
> 1x link with only <7% CPU usage (single CPU xeon 700MHz).
> I was very impressed.
>
> Go thru the socket layer API would just slow thing down.
>[...]
Do you have any data to back up this claim about socket layer?
Also, time and time again I see the same bullshit (sometimes
from intelligent people!). OF COURSE, it is TRIVIAL to send
messages (or receive messages), in fact, with a correct HCA
design IB allows it to be done from user mode. It is not where
the problem is. The problem is the delivery of notifications.
A notification is an interrupt (unless your application hogs
the CPU by CQ polling), and an interrupt is a context switch.
It takes infinitely smaller amount of time to post a work
request than to retire it on Infiniband. Best people are
cracking their heads over it. Ashok Raj, THE Intel's ARCHITECT
went as far as writing to linux-kernel asking for additional
arguments to RT signals (crazy, huh? :). Oracle gave up and said that
a) they do not mind hogging CPUs, they are the only application
running on the box anyways, and b) they only request notifications
for one in N messages [Joel, I know you are listening, care
to nod or call foul here?]
I may sound like Andre Hendrik here, but please understand
that nobody gives a flying dick about "sending 10k 32KBytes
message (320MB of data) every ~1 second over > 1x link with only
<7% CPU usage at a single CPU xeon 700MHz." My stack can do it today,
in 1500 lines of C code. It is the notification that counts.
How many RPC messages per second your clustered SQL server
can receive AND PROCESS is how many transactions it can do.
-- Pete
next prev parent reply other threads:[~2002-05-14 20:42 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-05-14 20:19 InfiniBand BOF @ LSM - topics of interest Tony.P.Lee
2002-05-14 20:42 ` Pete Zaitcev [this message]
2002-05-14 23:25 ` Jakob Østergaard
2002-05-14 23:58 ` Alan Cox
2002-05-15 15:45 ` Timothy D. Witham
-- strict thread matches above, loose matches on Subject: below --
2002-05-16 23:42 Woodruff, Robert J
2002-05-16 10:37 Eric Brunner-Williams in Portland Maine
2002-05-15 23:58 Woodruff, Robert J
2002-05-16 0:32 ` Alan Cox
2002-05-15 20:47 Tony.P.Lee
2002-05-15 21:36 ` Russell Leighton
2002-05-15 22:41 ` Lars Marowsky-Bree
2002-05-16 4:51 ` Ashok Raj
2002-05-15 20:37 Woodruff, Robert J
2002-05-15 20:55 ` Pete Zaitcev
2002-05-15 21:01 ` Alan Cox
2002-05-15 1:29 Tony.P.Lee
2002-05-15 2:35 ` Alan Cox
2002-05-15 5:01 ` Pete Zaitcev
2002-05-15 11:29 ` Alan Cox
2002-05-15 14:27 ` Ashok Raj
2002-05-15 16:34 ` Stephen Hemminger
2002-05-14 22:15 Woodruff, Robert J
2002-05-15 2:42 ` Timothy D. Witham
2002-05-13 16:25 Woodruff, Robert J
2002-05-13 16:36 ` Pete Zaitcev
2002-05-14 20:09 ` Timothy D. Witham
2002-05-14 20:23 ` Pete Zaitcev
2002-05-14 20:33 ` Timothy D. Witham
2002-05-14 21:01 ` Timothy D. Witham
2002-05-13 23:42 ` Alan Cox
2002-05-13 23:26 ` Lars Marowsky-Bree
2002-05-14 0:19 ` Alan Cox
2002-05-14 0:19 ` Joel Becker
2002-05-15 0:11 ` Eric Brunner-Williams in Portland Maine
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=20020514164245.C13781@devserv.devel.redhat.com \
--to=zaitcev@redhat.com \
--cc=Tony.P.Lee@nokia.com \
--cc=linux-kernel@vger.kernel.org \
/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