From mboxrd@z Thu Jan 1 00:00:00 1970 From: Theodore Ts'o Subject: Re: fsync stuck at jbd2_log_wait_commit on NVMe devices Date: Thu, 13 Aug 2015 09:09:53 -0400 Message-ID: <20150813130953.GA26095@thunk.org> References: <20150812183750.GB3373@thunk.org> <20150812223456.GC3373@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "linux-ext4@vger.kernel.org" To: Roy Yang Return-path: Received: from imap.thunk.org ([74.207.234.97]:56712 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751692AbbHMNJ4 (ORCPT ); Thu, 13 Aug 2015 09:09:56 -0400 Content-Disposition: inline In-Reply-To: Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Aug 12, 2015 at 04:34:17PM -0700, Roy Yang wrote: > Hi Ted,=A0 >=20 > =A0 =A0Appreciate your time, and will get the log from normal work lo= ad. >=20 > =A0 =A0BTW, the previous JBD2 log is from the node which is in bad st= ate.=A0 > =A0 Normal node finish within 3 seconds; while on this node, it takes= 10 seconds. Ah, you didn't say that. If this is a node in a bad state, and we're not seeing any real delays from the jbd2 logs, then it sounds very much like the problem is a hardware issue (or maybe an NVMe device driver issue). You might want to use blktrace to see how long the underlying I/O operations are taking to complete. - Ted -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html