A video demonstration on ForgeRock OpenAM deployment as a standalone instance in a Tomcat Server was posted earlier on my blog. For a production ready environment, it is important to have multiple instances of OpenAM running in a site. In the video that’s embedded below, you’ll get to see:
– An existing deployment of ForgeRock OpenAM in a Linux Container
– Installation & Configuration of a new instance of ForgeRock OpenAM in a separate Linux Container joining an existing OpenAM deployment
– Configuration of a OpenAM site that includes two OpenAM instances as mentioned above
– Installation and Configuration of HA Proxy in a separate Linux Container
– Demonstration of load balancing by HA Proxy to the back end OpenAM Servers
A great deal of information required for doing a demonstration on OpenAM HA environment and in turn the rules to be used by the HA Proxy was picked up from Mark Craig’s blog.
The illustration below might give an idea on the infrastructure used in the video demonstration. There are three Linux Containers in a Ubuntu 14.10 host, one of which has the first instance of OpenAM, the second one has another instance of OpenAM and the third one running HA Proxy to load balance the requests to two instances of OpenAM. The client requests go to the Linux Container running HA Proxy, from where the HA Proxy redirects the requests to either one of the OpenAM instances running in two different Linux Containers.
If you’ve got some some bit of idea looking at the illustration above, I can assure you, it’ll be clearer watching the demonstration below. A word of caution though: the infrastructure used here is only for demonstration purpose, for a very serious highly available environment, you may have to consider other virtualization technologies.
On this site, I’ve written another couple of posts around ForgeRock OpenIDM. If you’re not familiar with OpenIDM, I’d recommend reading/watching those (mentioned below), before viewing the video log embedded at the end of this post.
Video logs on the links above are quite detailed, but if you’ve not much time to spare to watch all of it, not to worry, the installation of OpenIDM and configuration of MySQL as its internal repository is quickly covered in the following screen-cast as well.
The below illustration might give you a rough idea about the infrastructure that I used to perform the demonstration on OpenIDM Cluster Configuration. I’ve a Host Operating System of Ubuntu 14.10 in which there are many Linux Containers. Two of LXCs named ‘my-openidm’ & ‘my-openidm2’ are used to install two separate instances of ForgeRock OpenIDM (say ‘node1’ & ‘node2’). A directory named ‘/software’ on the Host OS that has all required binaries is shared as ‘/source’ inside the Linux Containers. For brevity, I’ve included only the relevant LXCs in the illustration as follows:
Please use the embedded video for a quick reference. For a detailed study on how the OpenIDM works in a cluster, please refer to the ForgeRock documentation.
Talking of another High Availability facility in MySQL named Replication, I invite you to read an interesting blog post on self healing recovery in MySQL 5.6 Replication here. For a detailed list of courses on MySQL database access this url.