Interesting solution there. I hope it actually works and in theory it
should because all you are doing is correcting the RAID set. Obviously
doing anything more than that to those drives is fatal to that system.
As to moving the drives you can move them to and from the CEC if you
keep them under the right controllers. If you move them to a newer
controller that modifies the striping then yeah that's an issue. It has
nothing to do with 4 or 5 or 6 drive disk cages. So long as the drives are
connected to the same RAID card it matters not on which port or in which
slot they exist. So moving for example from a CEC with 8 drives to a
5095/0595 with 12 slots you can put them anywhere. Exception of course is
the load source disk with it's specific rules.
As you also suspect with SAS it's all different. There is no specific
spot for the load source but us traditionalists like to keep it on the
first drive in the drawer connected to the first port on the RAID card just
so we know where the bugger is.
To your questions to about SOP without having another system? Yeah that's
a tough one. Either you replace it before you power down, which you had no
opportunity to do since it was alive when you powered it down. Or you bonk
it to get it going again, which certainly doesn't always work. Or you have
a high enough function RAID card that will IPL with the drive dead. Or you
move the drives temporarily to another LPAR (which is much the same as
you're doing now but without the physical move.) Or you restore from
I hope it all works out well for yoU!
- Larry "DrFranken" Bolhuis
www.iDevCloud.com - Personal Development IBM i timeshare service.
www.iInTheCloud.com - Commercial IBM i Cloud Hosting.
On 8/22/2015 7:32 PM, Roberto José Etcheverry Romero wrote:
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
After doing somebody a favour and exchanging a 0595 for a 5095 (wanted to
have everyting in the rack instead of racked CEC and tower EXP) i ended up
running around due to a failed load source drive (drive died when system
spun down and the dust removed), failed to do D-Type IPL due to a missing
7212 (seems it never got used after the initial setup)
I brought the drives to my lab, put them into an installed machine and
found the dead drive. I'm now rebuilding before moving those disks (and
controller) back to the original machine.
Here is my question:
What would be the SOP for rebuilding that disk, without having an
additional expansion with the same kind of slots and installed LIC? (found
out that moving from 520 CEC to 5095 is possible but not backwards, if 6
disks are present on a single cage, it dies with XXXX9023)
I haven't found a "rebuild disk drive" in the D-Type IPL...
I haven't tried it yet but i think SAS disks are exempt from the whole
ordeal since the SAS Controllers no longer care in which slot the LS is
(which may mean that it distributes the info...)
In any case, if i can find another way great and if not, i already have
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives