Preparing For High Availability Environment

Component and Prerequisites
  • Nodes: In a high availability environment, multiple instances of Control Center Director can be connected, which are called nodes. Each node has Engine and Web server. To understand the components of a Control Center Director node, refer to IBM Control Center Director Overview.
  • Database: IBM Control Center Director supports native database solutions such MS SQL server, DB2, Oracle.

    Prerequisite: All the nodes must connect to a common Database. This will allow the user to extract data seamlessly even if one of the connected Control Center Director nodes go down.

  • Shared Location for Package Directory: A shared location must be defined to place the packages from where the Control Center Director nodes can detect the package and upgrade the servers.

    Prerequisite: The location of Package directory must be same for all the nodes. This will help the user to upgrade all the nodes in a sync.

    For Windows, create a shared location accessible by all nodes
    • Example: NFS Path: \\<IP_address/host>\<shared directory >
    For Linux/UNIX, create a shared mounted location
    • Example: NFS Path: <IP_address/host>:/<shared directory>. Mount the shared NFS path like, IP_address/host>:/<mount address>

    Refer Configuring Control Center Director to understand how to configure a package.

  • Load Balancer of web console: A Load balancer is a client added server.

    Prerequisite: Following are the perquisites for load balancer:
    • The server IP address/ host and port of all the nodes must be added in the Load Balancer configuration file.
    • From the RESTful API, navigate to REST APIs for New Install section and use /new-install/download-ca-public-pem API to set the SSL client certificate.
    • Enable CORS.
    • Set the following mandatory headers in the configuration file:
      proxy_set_header X-SSL-CERT  $ssl_client_escaped_cert;      
      proxy_set_header Host $http_host; 
      proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
  • Load Balancer of Event Processor (EP): In a high availability environment, the servers that are monitored by a failed EP can be temporarily reassigned to other EPs in the cluster. If multiple EPs are installed, the controller EP periodically checks to see whether any EPs in the cluster are down. If the controller EP detects that an EP is down, the controller EP temporarily distributes the servers that the EP monitors to other EPs.
  • Certificate Based Authentication : All the Control Center Director nodes and the load balancer must either configure the same set of certificates or the authorization of all the certificates must be the same. In case of change in certificates, the same must be duplicated on other nodes and Load Balancer. For certificate authentication configuration of a single node, refer Certificate-based Authentication.