Scaling

To run RIG in a more robust way you should use more nodes. To be able to form cluster from these nodes, check our deployment guide. Mentioned deployment guide also provides configuration for kubectl and helm (v2 and v3). RIG guarantees distribution of API requests, event stream events and SSE/WS events.

Even though it's possible to run RIG anywhere, we recommend to use Kubernetes.

# Increase number of nodes in Kubernetes cluster to 3
# After few seconds already running RIG will auto-discover new nodes and form the cluster, this means:
# - automatic synchronization of APIs across the nodes - notice logs with API synchronization
# - redistribution of Kafka/Kinesis partitions - notice logs with partitions re-balancing
# - blacklist state synchronization
# - SSE/WS synchronization
#
# From this point entire communication will be distributed between the nodes
kubectl scale deployment/reactive-interaction-gateway --replicas 3

# Start right away with 3 nodes using Helm template -- (assuming you are in the deployment directory)
helm repo add accenture https://accenture.github.io/reactive-interaction-gateway
# Helm v3
helm install --set replicaCount=3 rig accenture/reactive-interaction-gateway
# Helm v2
helm install --set replicaCount=3 --name=rig accenture/reactive-interaction-gateway