X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C73388.09723C62@onstor-exch02.onstor.net>; Mon, 8 Jan 2007 16:49:37 -0800
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-class: urn:content-classes:message
Subject: Clio/Lambo Upgrade Issue Status and Direction
Date: Mon, 8 Jan 2007 16:49:36 -0800
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E01FF82C4@onstor-exch02.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Clio/Lambo Upgrade Issue Status and Direction
thread-index: AccziAlWv1JLEWxnS6eDVC3T+LD+7A==
From: "Jay Michlin" <jay.michlin@onstor.com>
To: "Larry Scheer" <larry.scheer@onstor.com>,
	"Tim Gardner" <tim.gardner@onstor.com>,
	"Paul Hammer" <paul.hammer@onstor.com>,
	"Caeli Collins" <caeli.collins@onstor.com>,
	"Eric Barrett" <eric.barrett@onstor.com>,
	"Erik Palomba" <erik.palomba@onstor.com>,
	"Sandrine Boulanger" <sandrine.boulanger@onstor.com>
Cc: "Andy Sharp" <andy.sharp@onstor.com>,
	"Brian DeForest" <brian.deforest@onstor.com>,
	"Jerry Lopatin" <jerry.lopatin@onstor.com>

Hello all,

I've had the opportunity to discuss this install issue with a number of
you, and this email will bring everyone up to date and also explain next
steps.

STATUS
Larry has been working on the upgrade corruption issue essentially full
time, and has had help from Andy the past week. There is now reasonable
certainty that the problem is in the BSD install command. However, we
don't yet know definitively what triggers the problem, though we suspect
it has to do with memory use at least sometimes.

The eventual fix, which we are committing for *Delorean*, will be to the
install command, and may include our completely replacing the upgrade
process. It's a potentially complex project that we could not reasonably
commit for Lambo.

In parallel, we are now searching the online updates to BSD to see if
someone else has identified this problem over the past few years and
checked in a fix to install. If so, we'll know in the next day or two.
But even if we find such a fix, we are concerned that adequately testing
it may be beyond the scope of Lambo. So our tentative recommendation is
not to count on such a fix. If we do find one we will turn to QE and CS
for a decision.

At the same time, we're developing a script to serve as a temporary
workaround for use in Lambo. The intent is that the script be largely
transparent to customers. Even so, we think we ought to discuss the
script and any associated process with CS and QE. After all, if the
script unexpectedly breaks in the field, CS ought to know what it is and
how to deal with it.

The script also has an inherent problem that it cannot upgrade earlier
releases to Lambo, since it is not present on earlier releases.
Therefore, if we choose the script as our temporary solution, we will at
least require processes to implement it on earlier releases.

NEXT STEPS
By Thursday we expect to have such a script developed and ready. We also
expect to have completed research on any possible existing install fix
from the open source community. Once those are in place I expect to
invite those on the To: line of this message to meet and decide together
about how to proceed. I haven't scheduled the meeting yet pending
completion of the work I mentioned. But you can expect the meeting to be
either Thursday or Friday.

In the meantime, I'll keep you up to date on devlopments.

jay
