From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michael Tokarev Subject: Re: kvm guest: hrtimer: interrupt too slow Date: Sat, 10 Oct 2009 01:22:16 +0400 Message-ID: <4ACFA988.6070500@msgid.tls.msk.ru> References: <4AC207B1.7020901@msgid.tls.msk.ru> <20091003231205.GA15015@amt.cnet> <20091007231733.GG5903@nowhere> <20091008005456.GA10032@amt.cnet> <4ACD9AB9.3080803@msgid.tls.msk.ru> <4ACD9F68.10303@msgid.tls.msk.ru> <4ACDF1EF.3070005@msgid.tls.msk.ru> <20091008195232.GA8350@amt.cnet> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Thomas Gleixner , Frederic Weisbecker , kvm To: Marcelo Tosatti Return-path: Received: from isrv.corpit.ru ([81.13.33.159]:35010 "EHLO isrv.corpit.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760853AbZJIVWy (ORCPT ); Fri, 9 Oct 2009 17:22:54 -0400 In-Reply-To: <20091008195232.GA8350@amt.cnet> Sender: kvm-owner@vger.kernel.org List-ID: Marcelo Tosatti wrote: [snip] > Would be useful to collect sar (sar -B -b -u) output every one second > in both host/guest. You already mentioned load was low, but this should > give more details. Here we go: http://www.corpit.ru/mjt/hrtimer-interrupt-too-slow/ Two incindents - cases when hrtimer: interrupt is too slow were reported in the guest (with Marcelo's patch so that min_delta is increased to 50% only), happened at 11:44:48 and 21:46:56 (as shown in guest-dmesg file). For both, there's `sar -BWbd' output for a 2-minute interval (starting one minute before the delay and ending one minute after) from both host and guest. > Was there swapping going on? Not as far as I can see, and sar output agrees. /mjt