From owner-freebsd-hackers Thu Nov 6 19:54:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id TAA06181 for hackers-outgoing; Thu, 6 Nov 1997 19:54:17 -0800 (PST) (envelope-from owner-freebsd-hackers) Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id TAA06176 for ; Thu, 6 Nov 1997 19:54:15 -0800 (PST) (envelope-from marcs@znep.com) Received: from znep.com (uucp@localhost) by scanner.worldgate.com (8.8.7/8.8.7) with UUCP id UAA12818 for hackers@freebsd.org; Thu, 6 Nov 1997 20:54:09 -0700 (MST) Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id UAA01559 for ; Thu, 6 Nov 1997 20:58:57 -0700 (MST) Date: Thu, 6 Nov 1997 20:58:57 -0700 (MST) From: Marc Slemko Reply-To: Marc Slemko To: hackers@freebsd.org Subject: ufslk2 Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I have a news server that likes making innd hang in ufslk2 for periods of up to 30 seconds every time a sync is done. Is there any easy way to find out what file system (or even file) it is blocking on while the sync is happening? Most are mounted async. I have increased kern.update from the default of 30 to something higher because it does give a performance gain. Unfortunately, it results in longer periods of hang when it happens. Decreasing it back to 30 results hurts the server overall, although it does obviously reduce the time taken to sync each time (and therefore the time innd hangs in ufslk2) a lot, but it still adds up to a similar number. Yes, mounting async hurts this too but has a better overall impact. Any work being done on having the system be nicer when syncing if it has a large amount of data to sync? This is on 2.2-stable from a few months ago.