We recently decided to create a 2 node iPrint cluster. After reading
several docks and successfully completing testing, we went into production.
Everything ran great for a week, then the apbends began.

We noticed all the abends were in apache when it was rotating logs, so
after some extra digging we found a TID on the need to cluster apache
(http://support.novell.com/techcenter...alogID=2356083).
We dilegently followed this tid, and boom. Apache would not load in the
cluster name space, and everytime it tried the system performed a core
dump. How fun!

Since we only have 400 printers, we just gave up, and moved backed to a
non-clustered system. But in the name of academic curiosity, does anyone
have any ides?

Boxes were/are only running iPrint.

Cluster load statement:

nss /poolactivate=NDPS_POOL
mount NDPS VOLID=254
CLUSTER CVSBIND ADD NDPS_SERVER x.x.x.x
NUDP ADD NDPS_SERVER x.x.x.x
add secondary ipaddress x.x.x.x
search add volume:/scripts
clusterup
load broker /ipaddress=x.x.x.x .broker.print.services.ukans /allowdup
delay 10
load ndpsm .manager.print.services.ukans
/dbvolume=grouch_NDPS.services.ukans /dnsname=iprint.x.x.x