From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dave Jones Subject: Re: test processes are not all killed Date: Wed, 2 Aug 2017 23:22:43 -0400 Message-ID: <20170803032243.v776bgfuwzjmirko@codemonkey.org.uk> References: <20170801093813.nw3zj4nqb2jqpdtx@linux> <20170801153823.4z6nloqtnwnd3fe7@codemonkey.org.uk> <20170802030921.mwhtygus3oambllj@linux> <20170802123735.dmkice7c5slkpeqp@codemonkey.org.uk> <20170803031712.pgjuefn2bbs4wyx5@hqi2-mobl3.ccr.corp.intel.com> Mime-Version: 1.0 Return-path: Content-Disposition: inline In-Reply-To: <20170803031712.pgjuefn2bbs4wyx5@hqi2-mobl3.ccr.corp.intel.com> Sender: trinity-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Dai Xiang Cc: trinity@vger.kernel.org On Thu, Aug 03, 2017 at 11:17:12AM +0800, Dai Xiang wrote: > > Ah. You might just have a *lot* of dirty pages to write out. > > Does iotop show that journald is writing ? > > use iotop -o: > 111 be/4 root 0.00 B/s 0.00 B/s 0.00 % 99.99 % > [kworker/0:2] <== sometime kworker/1:1 or so on > sometimes more processes called kworker run, do not find trinity. You can try the ftrace stuff I suggested to trace what those kworker threads are doing, but it really sounds like it's just accumulated a ton of dirty pages to write out. > > Unless you're particularly focussed on stressing filesystems, you might > > want to skip the sync related syscalls (fsync,fdatasync,sync,syncfs) > > How can i skip those syscalls? use multiple -x arguments. This is common enough, that it might be worth adding a --no-sync argument that just does this for all relevant syscalls. Dave