Re: SSA RAID-5 replace disk



After the physical replacement, you have to really get rid of pdisk0.
(In that order! 1. Physical replacement, 2. Delete the definition of the
old pdisk0.) It is required that you NOT keep the old definition in the
database in a case like this. (That disk has failed and you're throwing
it out, so banish its memory.)

rmdev -l 'pdisk0' '-d' (or smitty with "keep def in DB...no")

Then all you have to do is run cfgmgr and it will find, recognize, and
define the new disk. It will likely come up first as a system disk, with
both a pdisk number and an hdisk number, which you don't want. You'll
have to go into smitty and change it to an Array Candidate (SSA Free)
disk (which will delete its hdisk number), and then add it to the array.
It's new disk number might or might not be pdisk0 - and this is
something you have no control over. (AIX systems have the right to
completely renumber all the pdisks and hdisks on you without warning, so
it's best to just be prepared for that to happen someday. I deal with it
by keeping track of serial numbers.)

After you run cfgmgr, you should go into diag's SSA tools and see what
your new disk's pdisk number and status is. This is most easily
discovered with Link Verification. Then from diag, there is a direct
link into smitty SSA Disk Arrays to add the pdisk to your RAID5 array
and finish the job.

(We still have about 300 SSA disk volumes in everyday use, in all kinds
of RAID configurations.)

Roger Deschner University of Illinois at Chicago rogerd@xxxxxxx


On Thu, 29 Nov 2007, -ray wrote:

Think back to the days of SSA haha...

I have a SSA RAID-5 array with no hot spare. pdisk0 died. I set it to
servicemode, then physically replaced it before doing an rmdev. Now I
can't get the new disk defined. It shows up as defined, but when i try to
mkdev i get the below error. Perhaps remnants of the old disk hanging out
somewhere? The diag link verification show a break in the loop where
pdisk0 is. The array is still up, just degraded. Any advice on how to
prceed is appreciated.

# lsdev -Cl pdisk*
pdisk0 Defined 40-60-P SSA Physical Disk Drive
pdisk1 Available 40-60-19D1-16-P SSA160 Physical Disk Drive
pdisk2 Available 20-58-19D1-12-P SSA160 Physical Disk Drive
pdisk3 Available 20-58-19D1-05-P SSA160 Physical Disk Drive
pdisk4 Available 40-60-19D1-01-P SSA160 Physical Disk Drive
pdisk5 Available 20-58-19D1-09-P SSA160 Physical Disk Drive
pdisk6 Available 20-58-19D1-08-P SSA160 Physical Disk Drive
pdisk7 Available 40-60-19D1-04-P SSA160 Physical Disk Drive
# mkdev -l pdisk0
Method error (/usr/lib/methods/cfgssapdisk):
0514-022 The specified connection is not valid.


ray
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Ray DeJean http://www.r-a-y.org
Systems Engineer Southeastern Louisiana University
IBM Certified Specialist AIX Administration, AIX Support
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=




Relevant Pages

  • Re: SSA RAID-5 replace disk
    ... I forgot the menus but make sure it is setup as " array candidate " for the disk type. ... That is one thing that used to bite a lot of folks.Although I would expect pdisk0 to be available ... please immediately delete it and notify sender at the above CSX ...
    (AIX-L)
  • Re: errpt not clear to me
    ... While the array can still be used, ... as a double disk failure. ... Select List Status Of All Defined SSA RAID Arrays. ... > Subject: Re: errpt not clear to me ...
    (AIX-L)
  • Re: HP EVA4000 / IBM DS4300 / EMC CX3-20/40
    ... Both EMC and EVA are great arrays and they will serve you well. ... disk array with the virtual raidsets on top. ... So, the system admin, and the DBAs had to create and manage lots of ...
    (comp.arch.storage)
  • Re: Need feedback on the A5200 storage array....
    ... they don't have the money for a big Hitachi array or a fast FC array with ... Use RAID5 on that kind of hardware. ... ten years or so) that had internal RAID5 controllers. ... I can't simply yank a disk and read its ...
    (comp.unix.solaris)
  • Re: RAID 5 corruption, RAID 1 more stable?
    ... corruption to either the RAID array itself or the file system. ... The disk array to suffer so many errors (for example disk errors ... There is nothing the disk array can do if the host is broken and ...
    (comp.arch.storage)