X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C8B53B.17673718@onstor-exch02.onstor.net>; Tue, 13 May 2008 13:51:18 -0700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-class: urn:content-classes:message
Subject: Time for vsvr failover
Date: Tue, 13 May 2008 13:51:17 -0700
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E09EE7F13@onstor-exch02.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Time for vsvr failover
Thread-Index: Aci1Oxci1qtfGpOqQMKkp/wYWR30Sg==
From: "Rich LaReau" <rich.lareau@onstor.com>
To: "dl-cstech" <dl-cstech@onstor.com>


Hi all,

We have an issue at LSI where they have seen a vsvr move taking more
than five minutes.  I know that if a vsvr has lots of volumes this can
take a while, as they unmount sequentially.  But we're also curious
about the nature of the number (and type?) of connections to the shares.
That is, do we have any data or experience on what factors influence a
manual vsvr move?  LSI has thousands of NFS connections to the shares,
many of which might themselves be dead.

The concern is the amount of downtime some users will experience, if
they get shut down first but have to wait for the whole process to
complete to get back online again.  Rebooting the node altogether moves
everything over pretty quick, but it's not always the case that you want
to move all the vsvrs at once.  (Might we consider a vsvr move flag that
does a "hard but fast" failover?)

Thanks,
Rich
