L3/L4 Load Balancing with the MapR Data Access Gateway

You can use haproxy for L3/L4 load balancing of clients that use the MapR Data Access Gateway. This topic describes how to install, configure, and run haproxy, and how to set your client connection string to connect to the load balancing service.

Determine the server where you want to run the load balancing service. The server must be reachable by the clients using the Data Access Gateway. It also must be able to connect to the Data Access Gateway.
  1. Install the haproxy service on the server you have identified:
    sudo yum install haproxy
    sudo add-apt-repository ppa:vbernat/haproxy-1.7
    sudo apt update
    sudo apt install haproxy
    sudo zypper install haproxy
  2. Configure the haproxy service by setting the following parameters in the configuration file at /etc/haproxy/haproxy.cfg:
    1. Create a frontend section with the following parameters:
      frontend <section_name>
         mode tcp
         bind *:<port_to_use_in_the_client_connection_string>
         default_backend <backend_section_name>
    2. Create a backend section with one server entry for each Data Access Gateway server:
      backend <backend_section_name>
         mode tcp
         server <DAG_server_name1> <DAG_server_host1>:<DAG_server_port1>
         server <DAG_server_name2> <DAG_server_host2>:<DAG_server_port2>
         server <DAG_server_nameN> <DAG_server_hostN>:<DAG_server_portN>

      The <backend_section_name> is the parameter you specified in Step 2a.

  3. Restart the haproxy service:
    sudo service haproxy restart 

Setting Your Client Connection String

Assume you have the following haproxy configuration settings and you have installed haproxy on node1.cluster.com:

frontend connection_input
  mode tcp
  bind *:8553
  default_backend maprdb_servers

backend maprdb_servers
  mode tcp
  server srv01 node1.cluster.com:5678
  server srv02 node2.cluster.com:5678

You can use the following client connection string with this sample configuration: