I have come over a somehow strange problem behaviour of our system.
We have a Cluster with the Portland CDK running (on the master node == FlexLM host). Additionally there are a few Workstations for the develeopers running PGI Compiler.
Somehow since a few weeks, the FlexLM is beeing shut down on the Cluster master node. I see the following in the log:
16:28:09 (lmgrd) SHUTDOWN request from root at node eule
16:28:09 (lmgrd) lmgrd will now shut down all the vendor daemons
16:28:09 (lmgrd) Shutting down pgroupd pid=20131 because of signal 15
16:28:09 (pgroupd) Shutdown requested from > email@example.com > IP=172.27.80.120
16:28:09 (pgroupd) daemon shutdown requested - shutting down
16:28:09 (pgroupd) IN: “pgi90-cdk-linux” haas@cauchy (SHUTDOWN)
16:28:09 (pgroupd) IN: “pgi-dbg” haas@cauchy (SHUTDOWN)
16:28:09 (pgroupd) IN: “pgi-dbg-gui” haas@cauchy (SHUTDOWN)
16:28:09 (lmgrd) Shut down FLEXnet pgroupd license server system on machine keam.gap.fzk.de
16:28:09 (lmgrd) EXITING DUE TO SIGNAL 15
The client euler (should actually be euler) / cauchy are development machines running Suse Linux 10.x and 11. They have PGI 9.0 installed. The license is made available by the modules packeg (evnironment settings manager)
prepend-path PATH /opt/pgi/linux86-64/9.0/bin
prepend-path LD_LIBRARY_PATH /opt/pgi/linux86-64/9.0/lib:/opt/pgi/linux86-64/9.0/libso
prepend-path MANPATH /opt/pgi/linux86-64/9.0/man
setenv LM_LICENSE_FILE 27000@keam
setenv PGI /opt/pgi
Does anybody have an idea about the reason for this behaviour ???
or any solutions??? (I am running nagios on the cluster master and I could integrate the flexlm into magion, but then the reason for the problem will still exist!)
Regards fro snowy Garmisch-Partenkirchen