3. Configuring the cluster
For the configuration, mostly TN10 has been used: http://kb.juniper.net/InfoCenter/index?page=content&id=TN10
While configuring the cluster, there is one bug in JunOS 11.2 to keep in mind. When you use T-SFP modules for your control links, and you enable the cluster with:
root@node0> set chassis cluster cluster-id {clusterID} node {nodeID}
Your control link might stay down. Even when the switch shows link is up. But in fact you’re missing a HA interface and the link lights will be off on the SRX side. I’ll get back to this later on.
First execute the following command on both nodes to enable the cluster.
@Node0:
root@node0> set chassis cluster cluster-id 1 node 0 reboot
@Node1:
root@node1> set chassis cluster cluster-id 1 node 1 reboot
After both nodes have rebooted, your base for the cluster is set. Your both control links should be up. Let’s check.
{primary:node0}
root@node0> show chassis cluster interfaces
Control link status: Up
Control interfaces:
Index Interface Status
0 em0 Up
1 em1 Up
Your fabric will be down, because we haven’t configured anything for that yet. Please referer to TN10 for the further steps. This because they might update those steps.
Getting back to the issue with JunOS 11.2, when you first set up the cluster (when you issue the ‘set chassis cluster’ command) and you’re using T-SFP modules for the control links, be sure to check that the HA interface are present. Check this with:
{primary:node0}
root@node0> show chassis ethernet-switch
node0:
-------------------------------------------------------
Displaying summary for switch 0
Link is good on GE port 3 connected to device: RE-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 7 connected to device: FPC1
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 20 connected to device: CPP-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 22 connected to device: HA-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Enabled
node1:
--------------------------------------------------------
Displaying summary for switch 0
Link is good on GE port 3 connected to device: RE-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 7 connected to device: FPC1
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 20 connected to device: CPP-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Disabled
Link is good on GE port 22 connected to device: HA-GigE
Speed is 1000Mb
Duplex is full
Autonegotiate is Enabled
When one or both of the HA-GigE interfaces are missing, you have two options to fix it:
1. Cold reboot the node. Just completely remove the power of the node and start it again.
2. Remove and reinsert the T-SFP module. When you do this, the link lights will be lit again.
This issue arises only when you first setup your cluster. Future reboots etc. won’t bring this issue back.
Note: This issue is only with T-SFP module. When you use SFP lasers, this problem does not occur.
Well, now we’ve our cluster fully operational and up and running!
{primary:node0}
root@node0> show chassis cluster status
Cluster ID: 1
Node Priority Status Preempt Manual failover
Redundancy group: 0 , Failover count: 1
node0 100 primary no no
node1 1 secondary no no
Redundancy group: 1 , Failover count: 1
node0 100 primary no no
node1 1 secondary no no
{primary:node0}
root@node0> show chassis cluster interfaces
Control link status: Up
Control interfaces:
Index Interface Status
0 em0 Up
1 em1 Up
Fabric link status: Up
Fabric interfaces:
Name Child-interface Status
fab0 ge-0/0/4 Up
fab0 ge-0/0/5 Up
fab1 ge-4/0/4 Up
fab1 ge-4/0/5 Up
Redundant-ethernet Information:
Name Status Redundancy-group
reth0 Up 1
reth1 Up 1
This is just a basic setup. Next page is about all the issues I encountered during my “quest”.