Discussion

4
Replies
1937
Views
CarlJ282 Member since 2016 1 post
Nielsen
Posted: August 3, 2016
Last activity: October 4, 2018
Closed

How can the Hazelcast advertised IP be modified

Pega 7.2 runs fine in Docker as a standalone instance. If additional instances are started failures occur at startup when Hazelcast attempts to create a cluster. This is due to Pega using the Docker container IP instead of the IP address of the host machine. I have overridden the Hazelzast using " <env name="cluster/hazelcast/interface" value="10.255.??.???"/>" in prconfig.xml. Pega Hazelcast tries to bind to the interface and fails since the interface is not available directly from the Docker container.

Is there a way to specify 2 IP addresses for the Hazelcast cluster? An address to bind to and an address for incomming communication from other Hazelcast servers.

Deploying Pega in Tomcat 8.

Here is more detail related to the need.

The team is trying to deploy Pega within a docker container. When Pega starts within a docker container it binds to the IP of the docker container. This is the expected behavior. When the Pega instance is persisted into the database for node clustering, the IP used is the internal IP address of the docker container (127.0.0.1). The need is to specify the IP address of the docker host machine so that other nodes in the cluster can communicate with other members in the cluster.

The host IP is not considered an available interface within the Docker Container.

The behavior is not really related to hazelcast. It is a Pega cluster member registration behavior where Pega uses the host ip that hazelcast binds to and this value needs to be overriden when deploying in a Docker container. Cluster members need to use the actual Host IP to rather than the internal IP of the Docker Container.

unresolved.

**Moderation Team has archived post**

This post has been archived for educational purposes. Contents and links will no longer be updated. If you have the same/similar question, please write a new post.

System Administration
Moderation Team has archived post
Share this page LinkedIn