In this article we explain the Forced manual failover, when you should do it, and key points to remember.
The Availability groups in MS SQL Server work as magic when it comes to failovers for securing data. In terms of availability groups, there are three different kinds of failovers that you can perform. These include an Automatic failover, which does not involve any data loss, a planned manual failover, this also does not involve any kind of data loss, and then comes the forced manual failover, which does involve some amount of data loss. It can be understood when a user will have to perform automatic failover, or even planned manual failover, but what about forced manual failover? When to perform it? And what exactly are the scenarios where we need to initiate a forced manual failover? All this and more will be discussed below.
Manual Forced Failover and the Scenario it should be used in
When you are using the AlwaysOn Availability Groups, there is a possibility that the primary and the secondary replicas are not completely synchronized, in this kind of a situation, a forced manual failover can help solve the problem. Read the example below to understand it better.
Let us assume that we have an Availability Group configuration consisting two sites, a primary site, which holds the synchronous commit replica and the secondary site that holds the asynchronous commit replica. Now from this given situation, take away the first site, and just let the secondary site remain. This is primarily what happens when a failover occurs. The primary site becomes unavailable, and the role of the primary site is the taken over by the secondary site. Once the primary site is active and back in action, you will then be required to perform a forced manual failover to the secondary site, so that they get back to playing their original roles.
A forced manual failover is mainly performed in situations where the goal is disaster recovery. One of the key points to keep in mind when performing a forced manual failover to unsynchronized secondary replica is that there are high chances of losing data. It is thus the kind of option that you should use only when you can immediately restore the service to Availability groups, thus reducing the amount of data that is lost.
Points to Consider while performing a Forced Manual Failover
Do consider possibility of data loss – When you are looking to implement a manual failover, a data loss scenario looms over your head. In case you are dealing with a database where financial transactions are recorded, it is prudent to completely avoid initiating this option. For all other lines of business applications, you need to keep the application support teams in know about your action. At times you may need to even consider a third party specialized mdf fix tool to handle the scenario.
Make use of the Secondary Replica – Forced manual failover is best performed over secondary replica, and not on the primary replica.
Resume Data Movement – Once both the replicas are back in action, make sure that they get back to performing their originally assigned roles.
Victor Simon is a data recovery expert in DataNumen, Inc., which is the world leader in data recovery technologies, including mdb fix and sql recovery software products. For more information visit https://www.datanumen.com/