From mboxrd@z Thu Jan 1 00:00:00 1970 From: Wols Lists Subject: Re: mdadm stuck at 0% reshape after grow Date: Thu, 7 Dec 2017 17:06:50 +0000 Message-ID: <5A29752A.9060007@youngman.org.uk> References: <1865221512489329@web5g.yandex.ru> <20171206104905.GA4383@metamorpher.de> <61c9e4bd-1605-5b17-80ce-c738b80b7058@turmel.org> <20171206160346.GA5806@metamorpher.de> <1b43be27-f21a-1fba-f983-01c5356a654d@turmel.org> <20171207135832.GA4858@metamorpher.de> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20171207135832.GA4858@metamorpher.de> Sender: linux-raid-owner@vger.kernel.org To: Andreas Klauer Cc: Linux-RAID List-Id: linux-raid.ids On 07/12/17 13:58, Andreas Klauer wrote: > On Wed, Dec 06, 2017 at 02:19:17PM -0600, Edward Kuns wrote: >> 1) If I have bad blocks lists configured, how do I safely remove them? > > --assemble with --update=no-bbl is safe, since it only removes if empty. > If not empty, likely you'll end up doing --update=force-no-bbl anyway. > >> # smartctl -l scterc,70,70 /dev/sdb ; echo $? >> smartctl 6.5 2016-05-07 r4318 [x86_64-linux-4.8.13-100.fc23.x86_64] >> (local build) >> Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org >> >> SCT Commands not supported >> >> 0 > > It'd be hilarious if the timeout FUD on this list came with advice that > didn't even do anything for most people, and nobody ever noticed... > > Unfortunately, it returns 4 here. And there are years old posts that > explicitely check for it returning 4, so this shouldn't be new at all. > > Perhaps it's an intermittent error specific to your smartctl version? > > You can just set the timeouts unconditionally, if you really want them. > Except bash is back-to-front. True is 0, anything else is false. So I'm guessing the above drive you've quoted DOES support erc, therefore it's returned 0 (true) to say everything's okay. Does your drive support erc? I guess not? So an error code of 4 is *correct*, and in the sameple script on the wiki will trigger the code that sets the *kernel* timeout to 180. My Barracudas return 4 ... Cheers, Wol