From owner-FreeBSD-users-jp@jp.freebsd.org  Tue Dec 16 09:16:42 1997
Received: by jaz.jp.freebsd.org (8.8.8+3.0Wbeta7/8.7.3) id JAA19588
	Tue, 16 Dec 1997 09:16:42 +0900 (JST)
Received: by jaz.jp.freebsd.org (8.8.8+3.0Wbeta7/8.7.3) with ESMTP id JAA19583
	for <FreeBSD-users-jp@jp.freebsd.org>; Tue, 16 Dec 1997 09:16:39 +0900 (JST)
Received: from smicec-smtp.smicec.co.jp (smicec-smtp [172.20.48.128]) by smicecgw1.smicec.co.jp (8.8.5/3.5Wpl7-smicecgw1-970804) with ESMTP id JAA24457 for <FreeBSD-users-jp@jp.freebsd.org>; Tue, 16 Dec 1997 09:17:44 +0900 (JST)
Received: from [172.20.49.164] ([172.20.49.164]) by smicec-smtp.smicec.co.jp (8.8.5/3.5Wpl7-smicec-smtp-971026) with SMTP id JAA00978 for <FreeBSD-users-jp@jp.freebsd.org>; Tue, 16 Dec 1997 09:19:31 +0900 (JST)
Message-Id: <199712160019.JAA00978@smicec-smtp.smicec.co.jp>
In-Reply-To: <19971216035156S.sakae@tky.3web.ne.jp>
X-Mailer:  Macintosh Eudora Pro Version 3.0.2Jr2
Mime-Version: 1.0
Content-Type: text/plain; charset="ISO-2022-JP"
Content-Transfer-Encoding: 7bit
Date: Tue, 16 Dec 1997 09:16:58 +0900
To: FreeBSD-users-jp@jp.freebsd.org
From: Bunpei Matoba <bm@partner.smicec.co.jp>
Reply-To: FreeBSD-users-jp@jp.freebsd.org
Precedence: bulk
X-Distribute: distribute [version 2.1 (Alpha) patchlevel=24]
X-Sequence: FreeBSD-users-jp 22371
Subject: [FreeBSD-users-jp 22371] Re: "Out of mbuf clusters" and System
 down!!
Errors-To: owner-FreeBSD-users-jp@jp.freebsd.org
Sender: owner-FreeBSD-users-jp@jp.freebsd.org

$B=;6b@)8f$NE*>l$G$9!%(J

At 3:51 AM +0900 97.12.16, Yoshiaki Sakae wrote:
> $B$G(J, $B$\$/$,J9$-$?$$$N$O2?$+$H$$$&$H(J, $B>e5-$N$h$&$J%5%C%+!<%W%m%0%i%`$N%5(J
> $B%s%W%k$rF0$+$7$F$$$k$H$-$K(J, $BFMA3(J OS $B$4$HMn$A$F$7$^$C$?$N$G$9!%>!<j$K%j(J
> $B%V!<%H$7$F(J fsck $B$r$J$s$H$+>h$j1[$($?$"$H(J /var/log/messages $B$r8+$k$H(J,
>
> Dec 16 03:09:20 piku /kernel: Out of mbuf clusters - increase maxusers!
>
> $B$H$J$C$F$$$^$7$?!%$^$"(J, maxusers $B$NCM$rA}$d$7$F%+!<%M%k$r:F%3%s%Q%$%k(J
> $B$9$l$PNI$5$=$&$J$N$OJ,$j$^$9$,(J, $B$3$l$/$i$$$N$3$H(J(?)$B$G(J OS $B$4$HMn$A$F$7(J
> $B$^$C$F$$$$$N$G$7$g$&$+(J? $B$H$$$&$3$H$G$9!%(J

$B3N$+$K%j%V!<%H$7$F$7$^$&$N$OLdBj$@$H;W$$$^$9!%(J
$B$=$N$3$H$N@'Hs$O$5$F$*$$$F!$BP:v$H$7$F!$;d$N7P83$G$O(J
maxusers$B$rA}$d$7$F4V@\E*$K(Jmbuf cluster$B$N?t$rA}$d$9$h$j(J
conf$B%U%!%$%k$G(J
option nmbclusters = 2048
$B$H$$$&$U$&$KD>@\;XDj$7$FBg$-$/$9$k$[$&$,M-8z$@$H46$8$F$$$^$9!%(J

2.2.1R$B$G$O(Jkernel$B$N%G%U%)!<%k%H%3%s%U%#%0%l!<%7%g%s$G$O(J
NMBCLUSTERS=512+16*MAXUSERS
$B$GDj5A$5$l$F$$$^$7$?!%(J
$B$3$l$G!$(JNMBCLUSTERS$B$rBg$-$/$7$h$&$H$9$k$H(J
MAXUSERS $B$r$+$J$jBg$-$/$7$J$$$H$$$1$J$$$N$G$9$,!$(J
MAXUSERS$B$OJL$NItJ,$N%5%$%:$K$b8z$$$F$/$k$N$G(J
$B%+!<%M%k$N%a%b%j%5%$%:$,Bg$-$/$J$j$9$.$F(J
$B;W$o$LI{:nMQ$,=P$F$7$^$&$3$H$,$"$j$^$9!%(J
$B;d$N>l9g$O!$(JDE$B$N%$!<%5%M%C%H%+!<%I$N%I%i%$%P!<$N(J
$BF0$-$,@dITD4$K$J$j$^$7$?!%(J

<$B;29M(JFreeBSD$B$N(JFAQ>
http://www.freebsd.org/FAQ/FAQ73.html#73
> 4.9. FreeBSD 2.0 panics with ``kmem_map too small!''
>
> Note The message may also be ``mb_map too small!''
>
> The panic indicates that the system ran out of virtual memory for network
> buffers (specifically, mbuf clusters). You can increase the amount of VM
> available for mbuf clusters by adding:
>
> options "NMBCLUSTERS=<n>"
>
> to your kernel config file, where <n> is a number in the range 512-4096,
> depending on the number of concurrent TCP connections you need to support.
> I'd recommend trying 2048 - this should get rid of the panic completely. You
> can monitor the number of mbuf clusters allocated/in use on the system with
> netstat -m.

--
Bunpei Matoba
Computer System Division
Sumitomo Metal Industries Control & Engineering Corp.
Amagasaki Japan
Phone: +81-6-418-4707    fax. +81-6-418-4764
Nifty-Serve: JAE00767   CompuServe: 72300,3551
http://www.nerimadors.or.jp/~bunpei/index.html


