From mboxrd@z Thu Jan 1 00:00:00 1970 From: Josh Durgin Subject: Re: osd down after adding OSDs Date: Tue, 09 Aug 2011 10:15:10 -0700 Message-ID: <4E416B1E.5040904@dreamhost.com> References: <4E407BB0.1020409@dreamhost.com> Mime-Version: 1.0 Content-Type: text/plain; charset=GB2312 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mail.hq.newdream.net ([66.33.206.127]:57483 "EHLO mail.hq.newdream.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753893Ab1HIRPK convert rfc822-to-8bit (ORCPT ); Tue, 9 Aug 2011 13:15:10 -0400 In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: huang jun Cc: ceph-devel On 08/08/2011 06:29 PM, huang jun wrote: > thanks=A3=ACJosh > Here the osd.10.log for the second condition. > and osd.15-tail.log for the first, because osd.15.log is about 130MB, > i'm not permitted to upload it, so i tail 70000 lines of it. > btw if the recovery statechart in GetInfo state, what will happened i= f > we recept a MOSDPGLog message, the statechart going to Crashed state? It depends on when the message was sent. If it was sent during a previous peering, it is ignored before it reaches the state machine by checking pg->old_peering_msg in OSD::handle_pg_log. Otherwise, if the log was sent after the current peering attempt started, the primary does enter the Crashed state. It's a bug that a lo= g was received in this case, since GetInfo is the first stage of peering, and no logs were requested yet. -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html