I inherited responsibility for an ADSM server with a separate policy
domain for each client. I have been trying to reduce the number of
policy domains to the number of distinct backup policies. I copied
a node's schedule from its current policy domain to a new domain. I
then used an UPDATE NODE command to change the policy domain the
node belonged to. I then used a DEFINE ASSOCIATION command to
associate the node with the copy of its traditional backup schedule.
The next day the scheduled backup did not take place. I went to the
client node and invoked the backup/archive client manually to connect
to the server (this is apparently needed in some cases to let the
central schedular figure out what the client's IP address is). The
next day the backup still did not take place. I had the administrator
of the client kill and restart the ADSM scheduler process, logged on
manually again, and went through a few other machinations that might
or might not have had any effect. The next day the node finally
resumed normal backups. I am still not sure which of the things I
did were actually necessary. Does anyone know what steps are necessary
to get scheduled backups running after the COPY SCHEDULE, UPDATE NODE,
DEFINE ASSOCIATION process described above?

Please don't suggest calling ADSM Support. I did that and found them
completely unhelpful. They ignored the UPDATE NODE command and pressured
me to perform exactly the same problem determination procedure one
would use if a node stopped doing backups with no apparent cause.