Date: Sun, 17 Feb 2002 19:09:49 +0100 From: Bernd Walter <ticso@cicely8.cicely.de> To: freebsd-alpha@freebsd.org Subject: runningbufspace deadlock problem on -stable Message-ID: <20020217180949.GD42870@cicely8.cicely.de>
next in thread | raw e-mail | index | archive | help
I have one -stable production PC164 machine with the problem that runningbufspace increases over time. The box is running cvsup.de.freebsd.org and therefor has some load on the disks. runningbufspace should measure the IO load and give preference for sync and read transactions over async transactions. With that problem the machine deadlocks after some time. The problem isn't reproduceable on my stable toybox (NoName) at home and difficult to debug on the box it happens to me because it's remote. Neither is the problem reproduceable on any i386 or alpha -current box. I recently noted that it also happens on axp0.freebsd.org: [5]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [6]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [7]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [8]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [9]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [10]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 [11]axp0> sysctl vfs.runningbufspace vfs.runningbufspace: 120832 The value should return to 0 when the box is IO idle. If someone sees this problems on a local machine I would be happy. I have some patches here adding assertions which I want to avoid trying on my production machine. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020217180949.GD42870>