In PRIMECLUSTER, a part of the processes work in a real time class. If you add or remove the system board using DR, the following message will appear:
Dec 25 21:12:41 Real time processes[pid= 4038 4218 4216 4286 4286 4286 4286 4046 4220 4134 4134 4134 4134 4134 4214 4221 4228 4287 4256 4291 4290 4288 4289 5350 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946 12946] is running. Do you continue DR ? [YES]/[NO]
Check if the displayed process IDs are for the PRIMECLUSTER daemons, then enter "yes" to continue with DR. (The same process ID will be repeatedly displayed for a multithread process.)
The real time processes that are started with PRIMECLUSTER are as follows:
PRIMECLUSTER
rcsd, bm, hvdet_*(Note 1)
(Note 1) The detector will be started up from the process bm as a real time process according to an application configuration. The process, in which the parent process is bm, and the process name starts with "hvdet_", is the PRIMECLUSTER detector.
If you frequently use DR, or if you want to make the DR process automatic, you can disable the message by creating an associated script and executing the "dr_conf" command. For details, see the "SPARC M10 System Domain Configuration Guide."
RMS might be suspended during DR. If this occurs, the following warning message might be output:
(SYS, 88): WARNING: No heartbeat from cluster host node0RMS within the last 10 seconds. This may be a temporary problem caused by high system load. RMS will react if this problem persists for 590 seconds more.
This message indicates that the RMS heartbeat monitoring is suspended temporarily. If this message appears during DR, you do not have to take any action.
