SUMMARY E3500 not seeing some of its internal drives

Hi all,

My thanks to:

Justin Stringfellow for confirming that probe-fcal-all would prove to be
beneficial.

William Enestvedt for providing some more information on what things are
supposed to look like and the following two links for drive information
http://docs.sun.com/db?p=/doc/805-2630-10/6j2h71m2l&a=view, and general
information about the 3500 http://docs.sun.com/db/doc/805-2630-10.

Vikas Khator for a suggestion (that drives get their power from the
fibre channel) that I was going to pursue after my musical chair dance
with fibre cables and plugs.

<sunsrv@blr.cmc.net.xx> for suggesting to check the FCAL loop.

probe-fcal-all showed both sbus-es, but only populated the lower plane.
I swapped the fibre patch cords around: no change.  I swapped the GBIC
plugin modules (or whatever they're called) on the IO board (between 0
and 1): no change.  I moved the GBIC modules from LA to LB and UA to UB,
and no change (well... the machine wouldn't boot any more, but I still
only saw the same disks).  I swapped the GBIC modules between LA and UA,
and now I saw the upper plane.

Note: if you're swapping fibre patch cords around, and probe-fcal-all,
you might end up with a hung machine (from the ok prompt, so not a BIG
big deal...).  Sometimes it might look like it's hung, but it returns
after a couple of minutes.  Sometimes moving a cable is not as
hot-swappable as you might like.  The secret is not to freak when all
your drives disappear after you've made a change...

Anyway: it seems like I need a new rinky-dink fibre module insert whammy
(or whatever they're called).

Apparently, Gert-Jan Hagenaars wrote:
% Hi all,
% 
% We have an E3500 at work, and it is only seeing the SCSI CDrom, and the
% bottom plane of (four) harddrives.  The drives in the top plane are not
% seen by the OS (they were before).  The box has not been moved, but it
% has been reinstalled (with Solaris 8).  Doing a "boot -r" from the ok
% prompt didn't make a difference.  Swapping the "unseen" drives to a spot
% in the lower plane does make them available, so the drives are not the
% problem.
% 
% I'm planning do to a "probe-fcal-all" from the ok prompt at my next
% maintenance window.
% 
% Supposedly there are _two_ fibre interfaces to the top plane for
% redundancy, but I don't know how to verify that.
% 
% I'm looking for suggestions on how to troubleshoot this thing (I have a
% screwdriver and a leatherman, and I'm not afraid to use them...)
% 
% CHeers,
% Gert-Jan.

CHeers,
Gert-Jan.

-- 
+++++++++++++ -------- +++++ --- ++ - +0+ + ++ +++ +++++ ++++++++ +++++++++++++
sed '/^[when][coders]/!d         G.J.W. Hagenaars -- gj at hagenaars dot com
    /^...[discover].$/d          Remembering Mike Carty 1968-1994
   /^..[real].[code]$/!d         UltrixIrixAIXHPUXSunOSLinuxBSD, nothing but nix
' /usr/dict/words                I'm Dutch, what's _your_ excuse?


CHeers,
Gert-Jan.

-- 
+++++++++++++ -------- +++++ --- ++ - +0+ + ++ +++ +++++ ++++++++ +++++++++++++
sed '/^[when][coders]/!d         G.J.W. Hagenaars -- gj at hagenaars dot com
    /^...[discover].$/d          Remembering Mike Carty 1968-1994
   /^..[real].[code]$/!d         UltrixIrixAIXHPUXSunOSLinuxBSD, nothing but nix
' /usr/dict/words                I'm Dutch, what's _your_ excuse?
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers