production ready?

production ready?

[Dwight Schauer]

Hello NILFS2 users and developers,

How stable do you consider NILFS2? Performance issues aside, I'm more
concerned about reliability.

I'm considering NILFS2 for a new server setup, as BTRFS does not
appear to production ready yet from what I've read.

So, what I'm asking is this, is NILFS2 considered production ready for servers?

Sincerely,
Dwight Schauer
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: production ready?

[Ryusuke Konishi]

Hi,
On Wed, 13 Oct 2010 12:14:46 -0500, Dwight Schauer wrote:
> Hello NILFS2 users and developers,
> 
> How stable do you consider NILFS2? Performance issues aside, I'm more
> concerned about reliability.
> 
> I'm considering NILFS2 for a new server setup, as BTRFS does not
> appear to production ready yet from what I've read.
> 
> So, what I'm asking is this, is NILFS2 considered production ready for servers?
> 
> Sincerely,
> Dwight Schauer

NILFS2 is almost stable.  We have a record of nine months operation on
in-house samba servers and a webDAV server.

However, a few crash problems are reported on this mailing list and
remain unresolved.  I suspect they are caused by some sort of bug
related to "garbage collection" mechanism, which often makes trouble
on this sort of filesystems.

I think you may as well wait for resolution of these issues and test
enough if you will actually apply it for business use.  NILFS2 is
still marked "experimental" in kernel as with BTRFS.

Thanks,
Ryusuke Konishi
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: production ready?

[Jérôme Poulin]

On Thu, Oct 14, 2010 at 1:50 AM, Ryusuke Konishi
<konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org> wrote:
> Hi,
...
> NILFS2 is almost stable.  We have a record of nine months operation on
> in-house samba servers and a webDAV server.
>
> On Wed, 13 Oct 2010 12:14:46 -0500, Dwight Schauer wrote:
>> Hello NILFS2 users and developers,
>>
>> How stable do you consider NILFS2? Performance issues aside, I'm more
>> concerned about reliability.
>>

On my side, it is on my laptop /home partition for a bit more than a
year, first month had corruption because of a bug which was fixed,
then a 2 months later corruption caused by garbage collector not
acting correctly when date is adjusted backward (in fact, forward then
backward in my case). But now it is corruption free and I can see it
when making a backup, no more I/O errors! Had some partition full
problems because garbage collector wouldn't start doing it's job until
I restarted it, but else it's OK.
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: production ready?

[Dwight Schauer]

On Thu, Oct 14, 2010 at 7:25 AM, Jérôme Poulin <jeromepoulin@gmail.com> wrote:
> On my side, it is on my laptop /home partition for a bit more than a year...

On Thu, Oct 14, 2010 at 1:50 AM, Ryusuke Konishi
<konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org> wrote:
> NILFS2 is almost stable.  We have a record of nine months operation on
> in-house samba servers and a webDAV server.

OK, Thanks.

There is still still warning when mounting verbosely though:
  mount.nilfs2: WARNING! - The NILFS on-disk format may change at any time.
  mount.nilfs2: WARNING! - Do not place critical data on a NILFS filesystem.

As per the first message, Is the on disk format expected to change any
time soon? In other words, what is the likelihood of my using NILFS
when 2.6.35.x, then having some future kernel upgrade render my NILFS
formatted filesystems un-mountable because the on-disk format changed?

I understand the disclaimer of the the second message.
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: production ready?

[Ryusuke Konishi]

On Thu, 14 Oct 2010 07:55:44 -0500, Dwight Schauer wrote:
> On Thu, Oct 14, 2010 at 7:25 AM, Jérôme Poulin <jeromepoulin@gmail.com> wrote:
> > On my side, it is on my laptop /home partition for a bit more than a year...
> 
> On Thu, Oct 14, 2010 at 1:50 AM, Ryusuke Konishi
> <konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org> wrote:
> > NILFS2 is almost stable.  We have a record of nine months operation on
> > in-house samba servers and a webDAV server.
> 
> OK, Thanks.
> 
> There is still still warning when mounting verbosely though:
>   mount.nilfs2: WARNING! - The NILFS on-disk format may change at any time.
>   mount.nilfs2: WARNING! - Do not place critical data on a NILFS filesystem.
> 
> As per the first message, Is the on disk format expected to change any
> time soon?

Well, I don't want to change disk format in a way that breaks
compatibility.  I'm considering to remove the above message at the
next utility release.

We still have potential to break compatibility to implement essential
features like extended attribute/posix ACLs.  However, I think
influential change should be carefully avoided or limited to the
minimum at this stage.

> In other words, what is the likelihood of my using NILFS
> when 2.6.35.x, then having some future kernel upgrade render my NILFS
> formatted filesystems un-mountable because the on-disk format changed?

I hope this never happens.  We have already started to use nilfs2 for
in-house systems.  My mention above is about forward-compatibility
(i.e. ability that older implementations can read the partition
generated by newer version), I won't break backward-compatibility
except for some extraordinary reason.

Thanks,
Ryusuke Konishi
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html