Over the coming weeks I will be posting a series of blog posts on common misconception, questions, issues etc. that I have run into over the years that we have been offering the Failover Engine. The most common question I do get asked is "what exactly is the difference between high availability vs. disaster recovery".
I will provide a more in depth explanation below, but the best and quickest way to remember this is:
- High Availability = LAN
- Disaster Recovery = WAN
Some groundwork before I jump into the more in depth explanation, the Failover Engine works in a Active-Passive setup, meaning only one server has the SolarWinds services started and running. The Failover Engine is not an Active-Active solution, meaning both servers have the SolarWinds services started and running. With that in mind, for this post I will refer to each server as the following
- Primary or Active server = SolarWinds services are active and running
- Secondary or Passive server = SolarWinds services are not active and running
As also illustrated below in the first image, the High Availability (HA) role is normally deployed in a LAN where communications are configured with the Public IP Address being shared by both the active/primary and passive/secondary servers. The active/primary server makes the Public IP visible and available to the network while the passive/secondary server leverages a packet filter installed by the Failover Engine on to hide the server and prevent network access since two machines with the same IP cannot be on the network at the same time.
In the event of a failure on the active/primary server the packet filter is removed from the passive/secondary server making it now assume the role as the active/primary server while simultaneously adding the packet filter to the server that was originally the active/primary server making it the passive/secondary server. Since both servers are sharing the Public IP address, DNS updating is not required.
When deployed in a Disaster Recovery role, the active/primary server and the passive/secondary server operates over a Wide Area Network (WAN) in different subnets. As a result, the active/primary and passive/secondary servers are configured with different Public IP addresses. In the event of a failover, the Failover Engine automatically updates DNS with the IP address of passive/secondary, so to the end users they just continue to access the SolarWinds server with the same DNS name they always use.
Any questions or comments, please ask in the comment section.