Showing results for 
Search instead for 
Did you mean: 

cluster did not start with cmnetd error

Trusted Contributor

cluster did not start with cmnetd error



Need your help in starting the cluster. My cluster failed with below error in syslog. Please tell me the reason for failure and solution to do it.


 cmclconfd[18669]: Unable to attach to DLPI: I/O error
 syslog: cmrunnode: Failed to validate the network configuration but will try to start the nodes anyway.
 cmclconfd[18671]: Request from root on node xip1pax to start the cluster on this node
 cmcld[18676]: Daemon Initialization - Maximum number of packages supported for this incarnation is 300.
 cmcld[18676]: Global Cluster Information:
 cmcld[18676]: Network Polling Interval is 2.00 seconds.
 cmcld[18676]: IO Timeout Extension is 0.00 seconds.
 cmcld[18676]: Auto Start Timeout is 600.00 seconds.
 cmcld[18676]: Failover Optimization is disabled.
 cmcld[18676]: Information Specific to node xip1pax:
 cmcld[18676]: Cluster lock disk: /dev/dsk/c6t0d1.
 cmcld[18676]: lan6  0x0025b3e8d0ba  bridged net:3
 cmcld[18676]: lan0  0x00248124c818  bridged net:1
 cmcld[18676]: lan4  0x002481240992  bridged net:2
 cmcld[18676]: lan3  0x0025b3e871fb    standby    bridged net:3
 cmcld[18676]: lan2  0x0025b3e871fa    standby    bridged net:2
 cmcld[18676]: Heartbeat Subnet:
 cmcld[18676]: Heartbeat Subnet:
 cmcld[18676]: Configured quorum disk(s) /dev/dsk/c6t0d1
 cmcld[18676]: Member Timeout is 189.12 seconds.
 cmcld[18676]: Max reformation duration is 227.95 seconds.
 cmcld[18676]: The maximum # of concurrent local connections to the daemon that will be supported is 1024.
 cmdisklockd[18679]: Changed to working directory /var/adm/cmcluster/cmdisklockd.
 cmdisklockd[18679]: cmdisklockd started
 cmdisklockd[18679]: added device: /dev/vg03xip:/dev/dsk/c6t0d1
 cmcld[18676]: Total allocated: 46081288 bytes, used: 3415680 bytes, unused 42665600 bytes
 cmcld[18676]: Cluster lock disk /dev/vg03xip:/dev/dsk/c6t0d1 is good
 cmserviced[18680]: Initializing
 cmserviced[18680]: Executing command: rm -f /var/adm/cmcluster/.cmserviced.*.socket
 cmserviced[18680]: Request to run service cmlogd
 cmserviced[18680]: Request to run service cmfileassistd
 cmserviced[18680]: Request to run service cmnetd
 cmnetd[18685]: Changed to working directory /var/adm/cmcluster/cmnetd.
 cmfileassistd[18684]: Changed to working directory /var/adm/cmcluster/cmfileassistd.
 cmnetd[18685]: Initializing
 cmnetd[18685]: Executing command: rm -f /var/adm/cmcluster/.cmnetd.*.socket
 cmnetd[18685]: Auto Failback is enabled.
 cmcld[18676]: Service cmnetd did not successfully start in 60 seconds
 cmcld[18676]: Exiting with status 1.
 cmdisklockd[18679]: cmdisklockd exiting
 cmserviced[18680]: Service cmnetd terminated due to a signal(9).
 cmclconfd[18665]: The cluster daemon aborted our connection (231).
 cmwbemd[2715]: The cluster daemon aborted our connection (231).
 cmclconfd[18671]: The Serviceguard daemon, cmcld[18676], exited with a status of 1.
 cmserviced[18680]: Unable to send 192 bytes (Broken pipe) (cl_local_cl_send fd 11).
 cmserviced[18680]: Unable to set status entry for service cmnetd.  Cluster Daemon aborted our connection.
 cmfileassistd[18684]: The cluster daemon aborted our connection (231).
 cmproxyd[2366]: The cluster daemon aborted our connection (231).





Stephen Doud
Honored Contributor

Re: cluster did not start with cmnetd error

Your syslog.log shows MEMBER_TIMEOUT is set to 189.12 seconds (odd number)


The CCAF contains this statement:

# The maximum value recommended for MEMBER_TIMEOUT is 60000000
# (60 seconds).


Suggest you change MEMBER_TIMEOUT to 60 seconds and see if cmnetd starts.

1- edit the cluster configuration ascii file (CCAF)

2- update MEMBER_TIMEOUT to 60 seconds (60000000)
3- cmapplyconf

4-try cluster start