From owner-freebsd-questions  Mon Jun 24 14: 3:44 2002
Delivered-To: freebsd-questions@freebsd.org
Received: from salmon.maths.tcd.ie (salmon.maths.tcd.ie [134.226.81.11])
	by hub.freebsd.org (Postfix) with SMTP id 75ACA37B47E
	for <freebsd-questions@FreeBSD.ORG>; Mon, 24 Jun 2002 14:02:10 -0700 (PDT)
Received: from walton.maths.tcd.ie by salmon.maths.tcd.ie with SMTP
          id <aa79934@salmon>; 24 Jun 2002 22:01:53 +0100 (BST)
To: sig <sig@uct.kiev.ua>
Cc: freebsd-questions@FreeBSD.ORG
Subject: Re: minfo output: junk in mbufs. can someone comment this? 
In-Reply-To: Your message of "Mon, 24 Jun 2002 19:56:53 +0300."
             <18021465385.20020624195653@uct.kiev.ua> 
Date: Mon, 24 Jun 2002 22:01:48 +0100
From: Ian Dowse <iedowse@maths.tcd.ie>
Message-ID:  <200206242201.aa79934@salmon.maths.tcd.ie>
Sender: owner-freebsd-questions@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-questions.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-questions>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-questions>
X-Loop: FreeBSD.ORG

In message <18021465385.20020624195653@uct.kiev.ua>, sig writes:
> 503/15000 pages used, 8048 mbuf-size slots
>Checking cluster ref counts
>Scanning cluster free list... 307 entries
>Scanning mbuf free list... 703 entries
>0xc18f1000: weird data 131331 @ 0x3 (min 0xc18f1014 max 0xc18f1100)
>0xc18f1100: weird data 2 @ 0x37 (min 0xc18f1114 max 0xc18f1200)
>0xc18f1200: weird data 2 @ 0x0 (min 0xc18f1214 max 0xc18f1300)
>0xc18f1300: weird data 0 @ 0x2 (min 0xc18f1314 max 0xc18f1400)
>0xc18f1400: weird data 137669068 @ 0x1 (min 0xc18f1414 max 0xc18f1500)
>0xc18f1500: weird data 30 @ 0x0 (min 0xc18f1514 max 0xc18f1600)
>0xc18f1600: weird data 0 @ 0x10 (min 0xc18f1614 max 0xc18f1700)
>Checking for unreferenced clusters.
>
>
>Is this a "normal" behavior?

Yes, a few  "weird data" lines are quite normal, because the mbuf
code allocates mbufs in groups of 16 at a time, and it does not
initialise the newly-allocated ones before putting them on the free
list.

However it is not normal for the number of mbufs to continuously
increase until they run out. If you leave the system for a few days,
does the number of mbufs and clusters reported by netstat -m continue
to rise? If so, check the "Send-Q" and "Recv-Q" columns of "netstat
-an" first to see if there is data in stuck connections. You can
also use minfo's "-x" option to dump all the mbuf contents to see
if you can see a pattern that might suggest what code is to blame.

Ian

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