I have had this problem for several different reasons.
1) Make sure that you are using the same network naming conventions in all places, i.e.: Short names vs. FQDN names in /etc/hosts on both the LPAR and the HMC
2) Make sure the /etc/resolv.conf is set up correctly, and that the nslookup returns the same as your entries in the host tables in #1 above. Once this is correct, I often then have to reboot the HMC. Once that's done, check to see if the IBM.* daemons are running on the LPAR. If not, do the following: /usr/sbin/rsct/bin/rmcctrl -z , then stopssrc -cs ctrmc, then /usr/sbin/rsct/bin/rmcctrl -A, then startsrc -s ctrmc and check for the IBM.* daemons running. If all but IBM.DRM and IBM.HostRM are running, start these manually with startsrc -s.
There may be an easier way to deal with this, but this is what has worked for me so far.