Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 7 Mar 2003 06:48:26 -0800
From:      Wes Peters <wes@softweyr.com>
To:        "Marc G. Fournier" <scrappy@hub.org>, Vallo Kallaste <kalts@estpak.ee>
Cc:        David Schultz <das@FreeBSD.ORG>, freebsd-stable@FreeBSD.ORG
Subject:   Re: "leak" in softupdates?
Message-ID:  <200303070648.26984.wes@softweyr.com>
In-Reply-To: <20030307081643.B15693@hub.org>
References:  <20030305204526.T38115@hub.org> <20030307101718.GA1908@kevad.internal> <20030307081643.B15693@hub.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Friday 07 March 2003 04:27, Marc G. Fournier wrote:
> On Fri, 7 Mar 2003, Vallo Kallaste wrote:
> > On Fri, Mar 07, 2003 at 01:00:33AM -0800, David Schultz
> > <das@FreeBSD.ORG> wrote:
> >
> > [...]
> >
> > > What I really don't understand is how you seem to manage to run
> > > into more bugs than everyone else.  (FYI, the libc_r/KVA_PAGES
> > > bug is now fixed in -STABLE.)
> >
> > He has rather unusual requirements compared to a lot of others. As I
> > understand he runs hundreds of jails and has thousands of processes,
> > putting real challenge to VM. All this means that FreeBSD
> > isn't ready to enterprise yet, no matter what advocacy has to say.
> > I'm glad Marc hasn't gave up and every now-and-then sends the
> > findings out to lists, for the benefit of others, particularly
> > developers. One word about "enterprise"; I'm sure Yahoo and similar
> > big players run also FreeBSD on "big boxes", but they have real
> > engineers in the field and have customised installations, I guess.
> > Off-the-shelf FreeBSD isn't ready yet.

Off-the-shelf FreeBSD is about as good as most commercial UNIX systems; 
you seem to have either overestimated their ability to handle loads that 
are absurd for the hardware or have had a much better experience than I 
have.

> As Vallo says above ... the 'bug' that Tor helped me fix this past
> week, with vnlru_proc, being a good example ... how many ppl are
> running their server with 132 active mount points?  From what I can
> tell, the bugs I'm hitting are all 'fringe bugs', stuff that you really
> have to be doing something extreme to hit ... but, as such, if I can
> get the bug fixed, its also one less bug that has the chance of hurting
> someone else ...

Yes, indeed, and I suspect bugs like that generally get fixed a lot faster 
here than when you submit a similar bug report in Solaris or HP-UX.

> As for 'the enterprise', I think the only thing that I really find
> about FreeBSD that 'hurts' is that I can't go above the 4Gig of RAM
> limit within ia32, even though there are boards that do support it ...

Being worked on.  Not so hard to do, much harder to do right.  Guess who 
took the easy sleazy path? ;^)  The other good news is that the intel 
network cards, both 10/100 (fxp) and 10/100/1000 (em) support 64-bit 
addressing, even in 32-bit PCI slots, so you'll have at least ONE enet 
interface that'll work reasonably fast.

-- 

        Where am I, and what am I doing in this handbasket?

Wes Peters                                               wes@softweyr.com


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200303070648.26984.wes>