My original question to sun-managers & sage-au summarises as
| a freshly unpacked A1000
[..]
| LUN 0 formats ok
| LUN 1 never says a word
| LUN 2 is claimed to have failed
[..]
| An error occured while performing the Health Check
Thanks to everyone that responded
Jason Harris
Andrew van der Stock
Simon Woodhead
Neil Glennon
M Gherman
Oscar Goosens
Imre Kolos
Geoff Lane
Scott Clark
Thomas Wardman
Lauradel Collins
Imre Kolos (Imre.Kolos@eth.ericsson.se) pointed me in the right
direction. Imre had a near identical experience, and
``It turned out the my sd.conf file was broken, actually, for
some unknown reason ... I answered `no' to the question of
pkgadd wether to replace existing file sd.conf''
Our admin file for JumpStart had the same effect as answering `no'.
I have since discovered the fault lies with raid manager 6.1.1
update 2's packaging. It had sd.conf edit scripts but did not treat
it as an editable file, so the machine had no hope of talking to non-
zero LUNs. Fixed the package by hand and we were up and running.
A bug report has been lodged with Sun, but 6.2.something is out so
it may be a non-issue there.
Other useful stuff to know when experiencing A1000 hassles include:
* avoid creating non-zero LUNs until LUN 0 is finished,
or better yet avoid parallel creation completely
(but managed to live dangerously & created the lot in one)
* avoid odd characters in module names
* Interrupting or mixing GUI/CLI tools is bad karma
* In general CLI control is preferred, but firmware
revision is safer with the GUI.
* A1000 Firmware may not be newest out of box
* Host OBP firmware level needs checking
* OS & osa patches important as always
* Editing vfstab at 2am does not result in a clean reboot :-)
-- Adam Barclay Senior Systems Consultant eBusiness Interactive adam@oz.org -> adam@ebinteractive.com.au
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:14:05 CDT