[SunHELP] Solaris 10 incremental upgrade vs larger jump
bjrichus at yahoo.com
Tue Sep 18 15:39:44 CDT 2012
I really like the way ZFS mirror disk pairs work.
We have a bunch of T2000
machines set up this way and can verify that in an emergency you only need to
pop the failed disk of the pair out, push in the replacement and with no more
than a couple of lines of simple command line input the thing is re-silvering
away and the end user population never knows anything is going on.
we only noticed the problem because of the yellow LED on the front of the
server that prompted the operations staff to call me. I then went to check the
logs and sure enough it was a failed device.
I'd not be overstating it to
say "It just works".
This is one thing that Sun got right (before Oracle
took over) and things took a turn in a different direction, but that's another
story entirely. :-)
> From: Richie
<listmail at triad.rr.com>
>To: sunhelp at sunhelp.org
>Sent: Tuesday, September
18, 2012 12:50 PM
>Subject: Re: [SunHELP] Solaris 10 incremental upgrade vs
>That is a very interesting response. Doing a fresh install is
on the table as well and replacing the old Disksuite mirror setups seems like
a great idea.
>Given the time of year I'm wondering if there is a Solaris 10
U11 just around the corner.
>On 9/18/2012 11:12 AM, Brian Richman wrote:
>> We have a number of machines (30+) that we will only use LU on to update
existing releases and will (as a matter of policy), do a fresh install when we
>> go from one update (i.e. U8) to another (i.e. U10).
>> As out
>> insist on minimizing disruption to end user processes but
>> sufficient systems to deliver that, this is the only time we get
to do any
>> re-work on machines so they reflect more modern techniques like
>> for boot/root when before everything was plain old UFS and
prayer that no disk
>> drives would ever fail, and update things like drivers
for Fiber HBAs and so
>SunHELP maillist - SunHELP at sunhelp.org
More information about the SunHELP