HP Serviceguard for Linux Version A.11.16 Release Notes, Third Edition, August 2006

Serviceguard for Linux Version A.11.16 Release Notes
Patches and Fixes in this Version
Chapter 152
cmmodnet command gets stuck in an infinite loop when zero subnet is
passed as an argument since that case was not being accounted for and
got stuck in a while(1) loop.
Resolution: During the parsing of the command args, catch this case and
error out right away with the message that non-zero arguments are
required for cmmodnet. Modify the cmmodnet parsing to catch the zero
subnet case and error out appropriately.
JAGaf17444: cluster cannot form using cmrunnode on boot
Problem: The Serviceguard RC script run at system boot time does not
properly retry to form the cluster for 600 seconds. Instead it gives up
after the first failed attempt to join the cluster. cmcluster.init does
not retry after it fails with output:
cmrunnode: Issuing cmrunnode again may succeed.
Resolution: Make sure cmcluster.init retries for 10 minutes if it fails
with that output.
JAGaf47433: Serviceguard locklund logs too much with default
logging levels
Problem: cmlocklund has excess information in the logs.
Resolution: Close the file descriptor when a resource is destroyed.
Defects Fixed in 11.16.07 and 11.16.08
The following defects have been fixed in HP Serviceguard for Linux
A.11.16.07 and A.11.16.08:
JAGaf30745: nmap on hacl-cfg/udp port causes inetd looping,
cmclconfd problem
Problem: When cmclconfd found a header for an invalid message, it
would exit without receiving the message. This left the message in
inetd’s socket buffer, causing inetd to spawn another cmclconfd pro-
cess. Serviceguard commands appeared to hang until they timed out
with an error message.
Resolution: cmclconfd now receives messages even if they are invalid.
See also JAGaf93186.