- HP Cluster Management Utility
- Java GUI may work:
java -jar cmugui.jar - pdsh is one of the commands from CMU
- sudo /opt/cmu/bin/cmu_power -p BOOT -n compnodeXX # boot a node
powering off compnode10 ...
spawning 1 task(s) ................
waiting for 1 task(s) ................ { last:compnode10 }
sleeping 10 seconds ................
powering on compnode10 ...
spawning 1 task(s) ................
waiting for 1 task(s) ................ { last:compnode10 }
/opt/cmu/bin/cmu_power finished
- sudo /opt/cmu/bin/cmu_monstat --nodes=compnode11 # show the list of sensors and values
- sudo /opt/cmu/bin/cmu_power -p OFF -a # Shutdown all nodes
- sudo /opt/cmu/bin/cmu_power -p OFF -n "compnode001 compnode002 ..." # Shutdown specific nodes
- for i in {81..90} ; do sudo /opt/cmu/bin/cmu_power -p OFF -n "compnode$i" ; done # Shutdown through loop
- for i in {81..90} ; do sudo /opt/cmu/bin/cmu_power -p STATUS -n "compnode$i" ; done # Check status through loop
- cmu_power -p REBOOT -n compcnodeXX # boot a node
- CMU configuration: /opt/cmu/etc/cmuserver.conf
- Ref: https://community.hpe.com/t5/Server-Clustering/NO-Netboot-reboot-not-working-CMU-7-2-ProLiant-SL4540-Gen8-RHES/td-p/6482202#.XkVtBOhKjuo
- Log file of each node from CMU command: /opt/cmu/tmp/ # Find *.output files
- /etc/fstab is the configuration
- sudo pdsh -w node# 'service netfs restart' # restarting nfs. Will not fix HW failure.
- ssh into the node and run
ip a. Check if ib card is detected. - ssh into the node and run
hostname -lto see if ib card is detected. - sudo pdsh -w node#
ibhoststo see if ib card is detected or working correctly. - If ib status is not correct, HW replacement might be necessary.