In our application, we see that the Gremlin Java Client does not respond when we send beyond 10 concurrent requests to AWS Neptune. When we submit more than 10 requests concurrently, the Completable future never completes. Below is the code snippet which creates the cluster

Cluster cluster = Cluster.build()
                .addContactPoint(clusterEndpoint).channelizer(SigV4WebSocketChannelizer.class)
                .port(clusterPort)
                .maxContentLength(CONTENT_LENGTH)
                .maxConnectionPoolSize(MAX_CONN_POOL_SIZE)
                .maxInProcessPerConnection(MAX_IN_PROCESS_PER_CONN)
                .maxSimultaneousUsagePerConnection(MAX_SIMUL_USAGE)
                .enableSsl(true).create();
Client client = cluster.connect();

where MAX_CONN_POOL_SIZE & MAX_IN_PROCESS_PER_CONN and MAX_IN_PROCESS_PER_CONN is set to 64

We send the gremlin script as a string to the client using the following java code

client.submit(query).all().get();

The above call never returns for the 11th request and any subsequent request has the same behavior. However if i create a new client to the same Neptune cluster the request succeeds till 10 concurrent requests are made. AWS Neptune cluster config is - db.r4.large Any idea why this behavior is observed?