Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 20 Jun 2001 01:09:38 -0700
From:      Terry Lambert <tlambert2@mindspring.com>
To:        "Ashutosh S. Rajekar" <asr@softhome.net>
Cc:        Matt Dillon <dillon@earth.backplane.com>, freebsd-hackers@FreeBSD.ORG
Subject:   Re: max kernel memory
Message-ID:  <3B305A42.781FC4DE@mindspring.com>
References:  <Pine.LNX.4.21.0106201112090.1235-100000@vangogh.indranetworks.com>

next in thread | previous in thread | raw e-mail | index | archive | help
"Ashutosh S. Rajekar" wrote:
> >     For the diskless case I don't know if you can make
> >     it to a million simultanious connections, but Terry
> >     has gotten his boxes to do a hundred thousand so we
> >     know that at least is doable.  But rather then spend a
> 
> Hmmm. I wonder how much TCP/IP tuning must have gone into getting this
> figure. Any comments Terry ?

I guess we beat you to the punch...

We have a product which is now shipping, and which currently
supports 1,000,000 concurrent connections.

As Matt implied about maxing out connections, you can make
it, if you are willing to work at it long enough.

I did several patches, which I've posted to -current or
-hackers or -stable, which are really very necessary;
not all of them have made it into FreeBSD, but they will
have to, if anyone else wants to duplicate the feat.  This
includes the maxfiles in /sys/conf/param.c, which everyone
said was gross, but no one could suggest a better way of
doing the job.  You already got the best one for free,
when I squashed the Heisenbug a while ago.

I could tell you the rest, but I'd have to kill you. ;-p


> >     huge amount of time trying to max out the connections
> >     you might want to consider distributing the connection
> >     load across a farm of front-end machines.

But Matt... 1,000,000 connections is cool.  When it rolls
over to that last one... 8-).


> But the fact is that Intel has poured money into our firm
> (along other investors), and this was to support this idea
> of a supposedly low-cost box. This box does have a
> proprietary PCI card that we are also building (the details
> are NDA'ed, my lips are sealed).

I guess you could tell me, but then you'd have to kill me...
;-p


> But going by the amount of customization required to be done
> before FreeBSD can be used as a server OS, how about starting
> an independant volunteer effort that would go into the details
> about providing a really heavy-duty FreeBSD server?

This is hard to do generally, with the way the FreeBSD
code is currently structured.  You would probably end up
with one-off configurations based on expected memory size
and role, which isn't really very useful (IMO).  People
have been pushing on this pretty hard lately, since there
was a recent benchmark debate the turned out to be nothing
more than bad tuning and a poor application architecture
match to the OS.


PS: You're ex-IBM, right?  I think I remember you from a
Linux FS project mailing list a while back... Me too... IBM
bought the startup I worked at about two years ago.  

-- Terry

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3B305A42.781FC4DE>