Possibly migrating from Solaris to Suse + ASM ???

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Your probably right. I remember a CPU failure about 1yr ago
and the system did crash ..
I was just reiterating what our sysadmin said.

Although, I will say that we were able to simply take the
CPU offline
and get the machine up and restart the database. Total
downtime was about 15 minutes
which is acceptable for us.

I don't think you can offline CPU's in Intel Linux servers?

- -peter


Alexei_Roudnev wrote:
> I had many CPU failures on SUN boxes. In 90% cases, no
matter what is the
> server (how much HA stuff embedded), system crashes.
> 
> And I never ever saw a single CPU failure on Intell
DELL 2850-th (saw
> recoverable memory failures).
> 
> BIG box (like Sun E4500) never can be reliable. Just
because it is BIG.
> Exceptions exists, but they have a skyrocketed price.
> 
> ----- Original Message ----- 
> From: "McAllister, Andrew"
<McAllisterAumsystem.edu>
> To: "Peter Santos" <psantoscheetahmail.com>; "Kevin Closson"
> <kevincpolyserve.com>
> Cc: <suse-oraclesuse.com>
> Sent: Thursday, January 04, 2007 7:01 AM
> Subject: RE: [suse-oracle] re: Possibly migrating from
Solaris to Suse + ASM
> ???
> 
> 
>> -----Original Message-----
>> From: Peter Santos [mailto:psantoscheetahmail.com]
> snip
>> My other concern was raised by our sysadmin. With
our
>> SUN machine, if one of the
>> CPU boards crashes, the machine will not always
crash
>> (not sure about this ..), but
> 
> 
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org


iD8DBQFFnV8goyy5QBCjoT0RAkc1AJ44gyha0bOwYTQnjp3zYlugl6kPnQCf
bLBu
o1XwBXSv7efq1vjIKZ0KnVY=
=xvIs
-----END PGP SIGNATURE-----

-- 
To unsubscribe, email: suse-oracle-unsubscribesuse.com
For additional commands, email: suse-oracle-helpsuse.com
Please see http://www.suse.com/oracl
e/ before posting

On Thu, 2007-01-04 at 15:10 -0500, Peter Santos wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Your probably right. I remember a CPU failure about 1yr
ago and the system did crash ..
> I was just reiterating what our sysadmin said.
> 
> Although, I will say that we were able to simply take
the CPU offline
> and get the machine up and restart the database. Total
downtime was about 15 minutes
> which is acceptable for us.
> 
> I don't think you can offline CPU's in Intel Linux
servers?

We have and old Dell 8450 (8-way, 900Mhz Xeon, almost 7
years old now)
and it had a CPU failure about 3 years ago.  The system did
crash, but
crash recovery rebooted the box, the system itself marked
the processor
offline and linux rebooted with 7 of the 8 CPU's.  Total
downtime was
about 15 minutes.

Of course, we had to schedule another outage window to
replace the
faulty CPU, which was another 15 minutes, but we basically
just lived on
one less processor for a few days.  The OS at the time was
RHEL 3.  It
was the hardware that offlined the CPU, not the OS, but the
OS was
totally fine with having only 7 CPU's instead of 8.

Later,
Tom



-- 
To unsubscribe, email: suse-oracle-unsubscribesuse.com
For additional commands, email: suse-oracle-helpsuse.com
Please see http://www.suse.com/oracl
e/ before posting