All API Gateways in a cluster must be configured with the same multicast address. This is good design, and a requirement for the API Gateway to start up successfully. The API Gateway checks for this at startup.
The API Gateway may contain the following line in the ssg_0_0.log file:
2014-10-10T15:24:07.141-0400 SEVERE 1 com.l7tech.server.cluster.ClusterBootProcess: At least two nodes in database have different multicast addresses com.l7tech.server.LifecycleException: At least two nodes in database have different multicast addresses
Resolving the issue
The following instructions listed below should be followed in order to troubleshoot and resolve the issue noted above. Once resolved, the API Gateway should start up as expected. If it does not start up as expected after following the instructions below, then a Support case should be opened up for further assistance.
- Run the following command from a root prompt: # mysql -e "use ssg; select * from cluster_info;"
- Verify there are different multicast addresses in the output from step one above. If so, then continue forward. If not, then there may be a different root cause of the startup issue, and it is then recommended to open a Support case for further assistance.
- Stop both database nodes in the cluster. It is very important that nothing is writing to the database at this time. Note that processing nodes are not involved in this solution. To stop a database node, run the following command from a root prompt: # service ssg stop
- Delete the contents of the cluster_info table. The contents of that table will be regenerated based on the API Gateway configuration after just a couple of minutes. To delete the contents of the cluster_info table, run the following command from a root prompt: # mysql -e "use ssg; delete from cluster_info;"
- Start only the primary database node and wait approximately two minutes. This will regenerate the cluster_info table contents which includes the multicast details. To start a database node, run the following command from a root prompt: # service ssg start
- Once step five is fully complete (cluster_info table has new content again), then start the secondary database node to complete the process. To start a database node, run the following command from a root prompt: # service ssg start