
to get a handle on what was going on... something like...
 - tcpdump -A -s 4096 -w /tmp/trace host 10.10.10.2 and port 80
 - fired up RAID Admin
 - did strings on /tmp/trace 

got cgi post/xml by replaying RAID Admin HTTP POST commands

getting it to work seems to require some specific set of eol chars

getting it to work requires hard tabs

got data via 
 cat PostTemplate  | sed "s/_PAGE_ID_/$i/" | nc 10.10.10.2 80 > Page$i

raid status info is on Page01

drives info is on Page16

decoded xml: wrote some little scripts to strip the xml, parse the xml
into a perl struct and then dump the perl struct with array indexes

the xml <array> does not refer to RAID--but rather sez an xml array follows

ACP-User: guest is the monitor user

ACP-Password: fkyaaa1i is the passwd public

Apple-Xsync: bottom refers to the bottom controller

status -6745 seems to mean that raid is powered down

passwords can be gotten from the com.apple.RAIDAdmin.plist file

xserve-raid-status exits status 1 means degraded rebuilding

xserve-raid-status exits status 2 means degraded w/ missing disk(s)

xserve-raid-status exits status 3 means DEAD

see head of xserve-raid-status for latest exit status values

add-member is usually 255, has value 254 just after failure? and 
is set to 0-100 when rebuilding.... at any rate, assume < 101 is
rebuilding

