Replication Between Clusters

This example demonstrates GemFire support for asynchronous replication between clusters.

View on GitHub

WAN replication allows remote GemFire clusters to automatically keep their region data consistent through the use of gateway senders and receivers. A gateway sender distributes region events to another, remote GemFire cluster. A gateway receiver configures a physical connection for receiving region events from gateway senders in remote GemFire clusters. The gateway senders and receivers can be configured in several different topologies based on specific business needs. For more information on example topologies and associated use cases see GemFire documentation on Multi-site WAN Configuration

In this example, two clusters are created on your local machine, each with a unique distributed system id and the WAN gateway configured for active-active, bidirectional region updates. The New York cluster (ny) has id=1 and the London cluster (ln) has id=2. Each cluster contains the same partitioned region (example-region) and each has parallel gateway senders, which means each server in the cluster will send data updates for the primary region buckets they hold. Alternately, you can configure serial gateway senders, where only one server in each cluster sends all data updates across the WAN. Serial gateway senders are typically used for replicated regions or when the order of events between different keys in a partitioned region needs to be preserved.

This example runs a single client that connects to the London cluster and puts 10 entries into the example-region and prints them. After the client app has run, both clusters will contain the data.

This example assumes that JDK11 and GemFire are installed.

Steps

  1. From the gemfire-examples/wan directory, build the client app example

     $ ../gradlew build
    
  2. Run the script that starts the London and New York clusters. Each cluster includes one locator and two servers. Each server configures one gateway sender, one gateway receiver and one partitioned region attached to the gateway sender.

     $ gfsh run --file=scripts/start.gfsh
    
  3. Run the client example app that connects to the London cluster and puts 10 entries into the example-region. The data will be automatically sent to the New York cluster, as well as printed to the console.

     $ ../gradlew run
    
  4. In one terminal, run a gfsh command, connect to the New York cluster, and verify the region contents

     $ gfsh
     ...
     Cluster-1 gfsh>connect --locator=localhost[10331]
     Cluster-1 gfsh>query --query="select e.key, e.value from /example-region.entries e"
     ...
    
  5. In another terminal, run a gfsh command, connect to the London cluster, and verify the region contents

     $ gfsh
     ...
     Cluster-2 gfsh>connect --locator=localhost[10332]
     Cluster-2 gfsh>query --query="select e.key, e.value from /example-region.entries e"
     ...
    
  6. Use other gfsh commands to learn statistics about the regions, gateway senders, and gateway receivers for each cluster.

     Cluster-1 gfsh>describe region --name=example-region
     Cluster-1 gfsh>list gateways
    
  7. In the terminal connected to the New York cluster, put another entry in the region and verify it is in the region on this cluster.

     Cluster-1 gfsh>put --key=20 --value="value20" --region=example-region
     Cluster-1 gfsh>query --query="select e.key, e.value from /example-region.entries e"
    
  8. In the terminal connected to the London cluster, verify the new entry has also been added to the region on this cluster.

    Cluster-2 gfsh>query --query="select e.key, e.value from /example-region.entries e"
    
  9. Exit gfsh in each terminal and shutdown the cluster using the stop.gfsh script

    $ gfsh run --file=scripts/stop.gfsh
    
  10. Clean up any generated directories and files.

    $ ../gradlew cleanServer