From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qk0-f178.google.com ([209.85.220.178]:33329 "EHLO mail-qk0-f178.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750756AbeDWE1B (ORCPT ); Mon, 23 Apr 2018 00:27:01 -0400 Received: by mail-qk0-f178.google.com with SMTP id c70so6046414qkg.0 for ; Sun, 22 Apr 2018 21:27:01 -0700 (PDT) From: Kirubakaran Kaliannan References: <20180413154313.GC32099@magnolia> In-Reply-To: <20180413154313.GC32099@magnolia> MIME-Version: 1.0 Date: Mon, 23 Apr 2018 09:56:58 +0530 Message-ID: <3bd0697599c117cfd4db6431a2cba270@mail.gmail.com> Subject: RE: statvfs() return large f_bavail on a 10TB drive Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: "Darrick J. Wong" Cc: linux-xfs@vger.kernel.org Hi Darrick, Yes this looks like the bitflip ! Do you have seen any occurrence of this before or maybe something related t= o a in-memory corruption ? Thaks -kiru -----Original Message----- From: Darrick J. Wong [mailto:darrick.wong@oracle.com] Sent: Friday, April 13, 2018 9:13 PM To: Kirubakaran Kaliannan Cc: linux-xfs@vger.kernel.org Subject: Re: statvfs() return large f_bavail on a 10TB drive On Fri, Apr 13, 2018 at 12:10:19PM +0530, Kirubakaran Kaliannan wrote: > Hi All, > > I have couple of xfs filesystem on a 10TB STAT drive. > > After 50% of filesystem is full, the df =E2=80=93kH is returning a large = f_bavail. > > > Filesystem Size Used Avail Use% Mounted on > /dev/dm-4 10T -12T 22T - /mnt/a > /dev/dm-3 10T -12T 22T - /mnt/b > > I tried statvfs() and got the same large f_bavail as well, > > total filesystem size =3D 9763506176 KB > total Available size =3D 21279656740 KB 21279656740KB =3D 0x100000000000 | (about 50% of the space) ? Bitflip? --D > total used size =3D -11516150564 KB > > We are using Ubuntu 3.18.19 Linux kernel/xfs. > > If I unmount and mount the filesysem back, the f_bavail values are back t= o > normal. > > Can you please point me, if this is something a known issue got fixed in > XFS in the later version, or any lead may help. > It is difficult at this point of time to upgrade the XFS for this specifi= c > issue. > > Thanks, > -kiru > -- > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html