From owner-freebsd-bugs Tue Feb 5 6:30:16 2002 Delivered-To: freebsd-bugs@hub.freebsd.org Received: from freefall.freebsd.org (freefall.FreeBSD.org [216.136.204.21]) by hub.freebsd.org (Postfix) with ESMTP id 3A39F37B41D for ; Tue, 5 Feb 2002 06:30:07 -0800 (PST) Received: (from gnats@localhost) by freefall.freebsd.org (8.11.6/8.11.6) id g15EU7S62151; Tue, 5 Feb 2002 06:30:07 -0800 (PST) (envelope-from gnats) Date: Tue, 5 Feb 2002 06:30:07 -0800 (PST) Message-Id: <200202051430.g15EU7S62151@freefall.freebsd.org> To: freebsd-bugs@FreeBSD.org Cc: From: Alan Judge Subject: Re: bin/32295 Reply-To: Alan Judge Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org The following reply was made to PR bin/32295; it has been noted by GNATS. From: Alan Judge To: freebsd-gnats-submit@freebsd.org Cc: lwa@teaser.fr Subject: Re: bin/32295 Date: Tue, 05 Feb 2002 14:27:48 +0000 We're seeing what appears to be exactly the same problem on 4.5-STABLE. I also see talk on the MySQL mailing lists from other FreeBSD users who appear to be hitting this problem. We're load testing a new MySQL-based system. Under some circumstances (not easy to repeat, but we're looking), the CPU of the server will go to 100% and stay there. It remains responsive enough to queries and goes back to normal on a restart. ktrace shows a steady loop with the timeout slowly counting down: 53446 mysqld 0.000009 CALL poll(0x8229000,0x3,0x1d52) 53446 mysqld 0.000009 RET poll 1 53446 mysqld 0.000009 CALL gettimeofday(0x81cfc8c,0) 53446 mysqld 0.000013 RET gettimeofday 0 53446 mysqld 0.000028 CALL poll(0x8229000,0x3,0x1d51) 53446 mysqld 0.000010 RET poll 1 53446 mysqld 0.000011 CALL gettimeofday(0x81cfc8c,0) 53446 mysqld 0.000013 RET gettimeofday 0 I'll do some more digging next time it happens and also use strace, since it seems to show more info. Since we're not in production yet, I can do some digging with a debugger when this next happens if anyone has suggestions as to where to look. Laurent, have you had any more info since you submitted this report? -- Alan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message