From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from 206-248-167-68.dsl.teksavvy.com ([206.248.167.68]:40695 "EHLO mail.isoar.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753743AbcGGWjl (ORCPT ); Thu, 7 Jul 2016 18:39:41 -0400 Message-ID: <577EDA03.9000308@isoar.ca> Date: Thu, 07 Jul 2016 18:38:59 -0400 From: "Andrew E. Mileski" MIME-Version: 1.0 To: Francesco Turco , linux-btrfs@vger.kernel.org Subject: Re: Frequent btrfs corruption on a USB flash drive References: <0120508a-b9e7-b9e7-4f27-79f982ee07fe@fastmail.fm> <577EC57C.8060607@isoar.ca> <0ad5fc23-9b10-5e95-888d-9717a8fb480d@fastmail.fm> In-Reply-To: <0ad5fc23-9b10-5e95-888d-9717a8fb480d@fastmail.fm> Content-Type: text/plain; charset=utf-8; format=flowed Sender: linux-btrfs-owner@vger.kernel.org List-ID: On 2016-07-07 17:13, Francesco Turco wrote: > > On 2016-07-07 23:11, Andrew E. Mileski wrote: >> How large is this USB flash device? > > 64 GB. > I don't know if there is an official recommended minimum size for btrfs, but I would expect 64 GB to be okay. I've personally set my minimum recommendation for btrfs at 120 GB based on my experience with failures in various flash devices from 4 to 30 GB. If you want to experiment, I have a theory that formatting single volumes with "-m single" can avoid a potential controller race in one specific situation, plus it helps to reduce the meta overhead on smaller devices. Lastly, the last two USB issues I investigated were both fixed by replacing the cables. Something to try if it is a cabled device. ~~ Andrew E. Mileski