From owner-freebsd-hackers Wed May 24 14:53:59 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id OAA25365 for hackers-outgoing; Wed, 24 May 1995 14:53:59 -0700 Received: from server.iadfw.net (root@server.iadfw.net [204.178.72.1]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id OAA25356 ; Wed, 24 May 1995 14:53:13 -0700 Received: (from jbryant@localhost) by server.iadfw.net (8.6.9/8.6.6) id QAA13381; Wed, 24 May 1995 16:52:35 -0500 From: Jim Bryant Message-Id: <199505242152.QAA13381@server.iadfw.net> Subject: mb_map full To: freebsd-hackers@FreeBSD.org Date: Wed, 24 May 1995 16:52:30 -0500 (CDT) Cc: davidg@FreeBSD.org, jbryant@endersbox.iadfw.net Reply-To: jbryant@endersbox.iadfw.net X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1119 Sender: hackers-owner@FreeBSD.org Precedence: bulk We are currently having a major problem. Our news server is a heavily used machine, is a pentium, has 128M RAM and 128M swap space, and is usually up to 80% swapbound. The problem is that the system seems to temporarily freeze for minutes at a time. Last night I noticed a syslog saying "mb_map full" around the time of the freeze, did a check, and found a load of those corresponding roughly with previous freeze-ups. Question: Would it be fixed by increasing MAX_KMAP in /usr/src/sys/vm/vm_map.h ? If not, then how can I eliminate this problem? Apparently, nothing is lost during the freeze-ups, and as a matter of fact, everything just picks up right where it left off. I have ruled out "swap_pager: out of space", as we only saw that again two days ago [time for yet another memory upgrade]. Jim -- All opinions expressed are mine, if you | "I will not be pushed, stamped, think otherwise, then go jump into turbid | briefed, debriefed, indexed, or radioactive waters and yell WAHOO !!! | numbered!" - #1, "The Prisoner" jbryant@server.iadfw.net, System administrator, Internet America