Error Message

Upon LLT startup (either at system boot, or upon manual restart / reconfig of the LLT driver), the following error message is observed:

LLT: configuring module...

LLT lltconfig ERROR V-14-2-15238 node 0 already exists in cluster X and has the address - XX:XX:XX:XX:XX:XX"

LLT lltconfig ERROR V-14-2-15664 LLT could not configure any link
 

Investigation of MAC addresses on the local system using "ifconfig -a" will find that the MAC address detailed in the error message does not exist to any local nodes of this cluster.

The MAC address detailed belongs to another node in another cluster in the networked environment.

The cluster ID that is identified in /etc/llttab on the cluster nodes is already in use.

To resolve the issue, modify the Cluster ID contained in /etc/llttab on all cluster nodes.

Example:

 # cat /etc/llttab
set-node node1
set-cluster 2
link eth3 eth-4c:b1:6c:91:31:a8 - ether - -
link eth4 eth-4c:b1:6c:91:31:a9 - ether - -

Change to:

 # cat /etc/llttab
set-node node1
set-cluster 19317      <<< modify this ID number to some other number in the range of 0-65535 
link eth3 eth-4c:b1:6c:91:31:a8 - ether - -
link eth4 eth-4c:b1:6c:91:31:a9 - ether - -

Note: The ID number must match on all nodes for the cluster to successfully join.

It is possible to determine if a certain Cluster ID number is in use elsewhere.
You may use the "lltdump" command for this purpose.

Syntax:

/opt/VRTSllt/lltdump -f (device) -V -P -c (cluster ID)

Syntax examples:

/opt/VRTSllt/lltdump -f /dev/bge:0 -V -P -c 128   (Solaris)

/opt/VRTSllt/lltdump -f eth3 -V -P -c 128   (Linux)

An example test of cluster ID 128 to determine if in use (Solaris):

bash-2.05# /opt/VRTSllt/lltdump -f /dev/bge:0 -V -P -c 128            <---- We are checking Cluster ID 128

AAK C 128 S 01 D 00 P 65535 rdy 80000081 seq 00655804 len 0060
   03 07 80 0A 00 00 00 01 FF FF 8C 00
AAK C 128 S 00 D 00 P 65535 rdy 80000081 seq 001fbc91 len 0060
   03 07 80 0A 00 00 00 00 FF FF 8C 00
^C
 

As packets are received, 128 is in use.

Where no packets are received when checking a specific Cluster ID that ID is either not in use or not visible to this local cluster.

Where the lltdump test is performed using a low-pri heartbeat link and no packets are detected it is likely the Cluster ID is not in use in the entire environment.
Therefore, that Cluster ID would be safe to use for the cluster.

Applies To

VCS 5.1 or later.