X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C8861B.668B622A@onstor-exch02.onstor.net>; Fri, 14 Mar 2008 14:36:02 -0700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-class: urn:content-classes:message
Subject: RE: mounting the secondary flash partition
Date: Fri, 14 Mar 2008 14:36:02 -0700
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E05C744D6@onstor-exch02.onstor.net>
In-Reply-To: <20080314143238.380b2f18@ripper.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: mounting the secondary flash partition
Thread-Index: AciGGu4FrTCJ/e1mQ4GRMO0ZKvV4IgAABsuA
From: "Sandrine Boulanger" <sandrine.boulanger@onstor.com>
To: "Andy Sharp" <andy.sharp@onstor.com>

It runs by itself on my machine after power failures sometimes. That's
what hppens on bsd flashes. The filer crashes (unfortunately too often),
and sometimes when it boots, fsck runs and delete files. We've seen too
many times in QA but nobody could ever find a resolution for this.
We'll do similar test with Linux: keep crashing FP and see how the fs on
flash does at reboot.

-----Original Message-----
From: Andy Sharp=20
Sent: Friday, March 14, 2008 2:33 PM
To: Paul Hammer
Cc: Eric Barrett; Larry Scheer; John Keiffer; dl-File System; dl-QA
Subject: Re: mounting the secondary flash partition

On OpenBSD, fsck is demonized, but the real problem is the filesystem
is corrupt.  fsck is just doing its job.

On Linux, it's a completely different situation and I don't expect that
fsck should normally ever be used unless something seriously bad is
wrong with the filesystem on the flash, in which case it should
probably be initialized anyway.  In the field I expect this to be so
rare as to be never seen.  Put it this way: how many engineers here at
Onstor have had to run fsck on their workstations?  Except for Jim, I'm
guessing damn close to zero  ~:^)

On Fri, 14 Mar 2008 11:05:22 -0700 "Paul Hammer"
<paul.hammer@onstor.com> wrote:

> Are there any instances where 2efsck will run automatically?=20
> =20
> Want to know if we think we are going to see the same type of
> failures in the field that we see today when on the bobcat flash fsck
> blows away files. Thanks,
> =20
> -Paul
>=20
> ________________________________
>=20
> From: Eric Barrett
> Sent: Fri 3/14/2008 10:58 AM
> To: Larry Scheer; John Keiffer; dl-File System; Andy Sharp
> Cc: dl-QA
> Subject: RE: mounting the secondary flash partition
>=20
>=20
> We should fix this before Cougar ships, maybe with the tune2fs
> command Raj linked.  We have lots of Linux-savvy customers and they
> will see this when they poke around, and open support cases on it.
> ("How do I run e2fsck on the flash?"  "Why aren't you fscking your
> flash regularly?"  Etc.)=20
>=20
> ________________________________
>=20
> From: Larry Scheer=20
> Sent: Friday, March 14, 2008 10:55 AM
> To: John Keiffer; dl-File System; Andy Sharp
> Cc: dl-QA
> Subject: RE: mounting the secondary flash partition
>=20
>=20
>=20
> John,=20
>=20
>    This is just the behavior of the ext3 file system used in Linux.
> After a certain number of reboots this message appears. It is ok to
> ignore this message.
>=20
> =20
>=20
> Andy might have more to say about it.
>=20
> =20
>=20
> Larry
>=20
> =20
>=20
> ________________________________
>=20
> From: John Keiffer=20
> Sent: Friday, March 14, 2008 8:27 AM
> To: Larry Scheer; dl-File System
> Cc: dl-QA
> Subject: mounting the secondary flash partition
>=20
> =20
>=20
> I notice the following 4 lines when I mount the secondary flash's
> root partition. Are these normal and expected (the second one seems
> the most concerning)?
>=20
> =20
>=20
> Mar 14 08:21:01 g6r10 kernel: kjournald starting.  Commit interval 5
> seconds
>=20
> Mar 14 08:21:01 g6r10 kernel: EXT3-fs warning: maximal mount count
> reached, running e2fsck is recommended
>=20
> Mar 14 08:21:01 g6r10 kernel: EXT3 FS on sda1, internal journal
>=20
> Mar 14 08:21:01 g6r10 kernel: EXT3-fs: mounted filesystem with
> ordered data mode.
>=20
> =20
>=20
> Thank you,
>=20
> John Keiffer
>=20
> =20
>=20
> =20
>=20
