From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cloud1-vm154.de-nserver.de ([178.250.10.56]:34258 "EHLO cloud1-vm154.de-nserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752213AbcI2KFk (ORCPT ); Thu, 29 Sep 2016 06:05:40 -0400 Subject: Re: BTRFS: space_info 4 has 18446742286429913088 free, is not full To: Adam Borowski References: <1c94f3b1-b238-6668-7976-c0594f170dfe@profihost.ag> <57EBAAF5.10509@cn.fujitsu.com> <4f91af54-78ae-5c08-3daa-8a0c16210e5f@profihost.ag> <57EBB329.9060009@cn.fujitsu.com> <3e429001-020a-748b-c0c5-85b4091599a1@profihost.ag> <57ECBAE7.5030807@cn.fujitsu.com> <0e556b63-7c85-1e97-00b0-91bfca44e82c@profihost.ag> <57ECBF0D.3050609@cn.fujitsu.com> <85f85ca6-95c0-1394-9b09-93dcc00c2319@profihost.ag> <20160929100311.GA6357@angband.pl> Cc: "linux-btrfs@vger.kernel.org" From: Stefan Priebe - Profihost AG Message-ID: <8875e043-cada-2376-53e4-d14d2bf9c78a@profihost.ag> Date: Thu, 29 Sep 2016 12:05:36 +0200 MIME-Version: 1.0 In-Reply-To: <20160929100311.GA6357@angband.pl> Content-Type: text/plain; charset=windows-1252 Sender: linux-btrfs-owner@vger.kernel.org List-ID: Hi, Am 29.09.2016 um 12:03 schrieb Adam Borowski: > On Thu, Sep 29, 2016 at 09:27:01AM +0200, Stefan Priebe - Profihost AG wrote: >> Am 29.09.2016 um 09:13 schrieb Wang Xiaoguang: >>>>> I found that compress sometime report ENOSPC error even in 4.8-rc8, >>>>> currently >>>> I cannot confirm that as i do not have anough space to test this without >>>> compression ;-( But yes i've compression enabled. >>> I might not get you, my poor english :) >>> You mean that you only get ENOSPC error when compression is enabled? >>> >>> And when compression is not enabled, you do not get ENOSPC error? >> >> I can't tell you. I cannot test with compression not enabled. I do not >> have anough free space on this disk. > > Disabling compression doesn't immediately require any space -- it affects > only newly written data. What you already have remains in the old > compression setting, unless you defrag everything (a side effect of > defragging is switching existing extents to the new compression mode). Yes i know that but most workload is creating reflinks to old files and modify data in them. So to create a good test i need to defrag and uncompress all those files. Greets, Stefan