| From: |
| Andres Freund <andres-AT-anarazel.de> |
| To: |
| Jeff Layton <jlayton-AT-redhat.com> |
| Subject: |
| Re: fsync() errors is unsafe and risks data loss |
| Date: |
| Fri, 13 Apr 2018 06:25:35 -0700 |
| Message-ID: |
| <20180413132535.6o3ijzmk6birmvay@alap3.anarazel.de> |
| Cc: |
| Dave Chinner <david-AT-fromorbit.com>, lsf-pc <lsf-pc-AT-lists.linuxfoundation.org>, Matthew Wilcox <willy-AT-infradead.org>, Andreas Dilger <adilger-AT-dilger.ca>, "Theodore Y. Ts'o" <tytso-AT-mit.edu>, Ext4 Developers List <linux-ext4-AT-vger.kernel.org>, Linux FS Devel <linux-fsdevel-AT-vger.kernel.org>, "Joshua D. Drake" <jd-AT-commandprompt.com> |
Hi,
On 2018-04-13 09:18:56 -0400, Jeff Layton wrote:
> Yes, I think we ought to probably do the same thing globally. It's nice
> to know that xfs has already been doing this. That makes me feel better
> about making this behavior the gold standard for Linux filesystems.
>
> So to summarize, at this point in the discussion, I think we want to
> consider doing the following:
>
> * better reporting from syncfs (report an error when even one inode
> failed to be written back since last syncfs call). We'll probably
> implement this via a per-sb errseq_t in some fashion, though there are
> some implementation issues to work out.
>
> * invalidate or clear uptodate flag on pages that experience writeback
> errors, across filesystems. Encourage this as standard behavior for
> filesystems and maybe add helpers to make it easier to do this.
>
> Did I miss anything? Would that be enough to help the Pg usecase?
>
> I don't see us ever being able to reasonably support its current
> expectation that writeback errors will be seen on fd's that were opened
> after the error occurred. That's a really thorny problem from an object
> lifetime perspective.
It's not perfect, but I think the amount of hacky OS specific code
should be acceptable. And it does allow for a wrapper tool that can be
used around backup restores etc to syncfs all the necessary filesystems.
Let me mull with others for a bit.
Greetings,
Andres Freund
