From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-it0-f50.google.com ([209.85.214.50]:44530 "EHLO mail-it0-f50.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751668AbdIKMtW (ORCPT ); Mon, 11 Sep 2017 08:49:22 -0400 Received: by mail-it0-f50.google.com with SMTP id r131so6890002itc.1 for ; Mon, 11 Sep 2017 05:49:22 -0700 (PDT) Subject: Re: Please help with exact actions for raid1 hot-swap To: Marat Khalili , Patrik Lundquist Cc: "linux-btrfs@vger.kernel.org" References: <01258ddc-1d34-4cc9-d668-21cbe88ab1d0@rqc.ru> <9D2B6B65-7DE9-4922-82DC-08BF3A3A2E09@rqc.ru> From: "Austin S. Hemmelgarn" Message-ID: Date: Mon, 11 Sep 2017 08:49:16 -0400 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Sender: linux-btrfs-owner@vger.kernel.org List-ID: On 2017-09-10 02:33, Marat Khalili wrote: > It doesn't need replaced disk to be readable, right? Then what prevents same procedure to work without a spare bay? > In theory, nothing. In practice, there are reliability issues with mounting a filesystem degraded (and you should be avoiding running any array degraded, regardless of if it's BTRFS or actual RAID (be that LVM, MD, or hardware)). It's also significantly faster to do it with a spare drive bay because that will just read from the device being replaced and copy data directly, while pulling the device to be replaced requires rebuilding the data (there is more involved than just copying, even with a raid1 profile).