When a cluster comes from the headless state, osafmsgnd on payload is running with high CPU usage.
Running 'top':
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
3537 root 20 0 199828 20884 20392 R 98,0 1,0 0:53.96 osafmsgnd
To reproduce the problem is enough to run one controller and one payload:
1. Start SC-1
2. Start PL-3
3. When the sync is done, stop SC-1
4. Start SC-1
5. Run 'top' on PL-3
The MSG problem is detected on the default branch, but it can exist on earlier versions.
strace on osafmsgnd process shows that osafmsgnd is in poll loop with closed file descriptor (17):
poll([{fd=21, events=POLLIN}, {fd=15, events=POLLIN}, {fd=17, events=POLLIN}, {fd=12, events=POLLIN}, {fd=19, events=POLLIN}], 5, 4294967295) = 1 ([{fd=17, revents=POLLNVAL}])
My guess that it's CLM file descriptor.
MSG also reported to syslog:
Feb 10 14:44:19 PL-3 osafmsgnd[3749]: ER saClmDispatch Failed with error 9
commit c3324ac95b353ad063234df26c9a2ef555724f47 (ticket-2299)
Author: Alex Jones ajones@rbbn.com
Date: Thu Sep 12 08:27:22 2019 -0400