HA and Clusters

During the lockdown, almost 90 percent of the IT staff is working from home and to work from home you need one important thing which is internet. Now imagine that you have a very important video call which of course is going to happen via internet and as soon as the call starts your home internet goes down and you no longer have access to the internet and people are waiting for you on the call as you are the host, well its nothing more than a nightmare to me or for any IT individual working from home, so what’s the solution?? Have at least 2 internet connections at home "nowadays you can get high speed broadband at a very low price "Thanks to JIO to create that kind of market", so what happens is that if one ISP goes down you have the other internet to fall back and get your work done

This concept of having 2 Instances of any service available to avoid any downtimes (no service availability) is called "High availability" in the IT world, in fact any business application you access anywhere has minimum 2 instances hosting that application for you in the background, this number "2" is the key when designing any IT solution as it provides redundancy

If you have an application which is a critical application and should be available 24/7, then you should definitely be hosting this application on more than 1 servers, it’s not necessary to have only 2 servers, you can have more than 2 as well and all these hosts (servers) form a group which we call in IT terms as a “Cluster”. Any group of servers hosting the same application is called as a cluster

You can form clusters at the Operating system level such as Linux clusters, Windows failover cluster or the database levels such as SQL Clustering or Oracle RAC (Real Application Cluster)

You can have application based cluster as well where the configuration of the cluster happens at the application level

In an ideal Enterprise, Clustering is maintained at all levels, you may have clustering at Network level, then OS level, Application level and database level

These clusters eliminate the single point of failure and provide High availability of the applications to the End users so that the hosted services are never unavailable