In this section, I m going to explain how to make a disaster recovery plan in your company
Most of the companies are moving their IT infrastructure to cloud in order to take advantage in terms of cost, scalability and flexibility. When you move your IT infrastructure to the cloud, there are different things needs to be considered. One of the most important thing is a disaster recovery strategy. The point is to note that if any disaster happens, you need to make sure the business process keeps going.
These are the basic disaster recovery strategies you need to consider;
- Backup strategy
- Pilot light
- Warm standby
Before selecting your disaster recovery strategy, it would be better to have an idea about them. For the different disaster recovery strategy, you need to put different budget and effort. If your company loses millions of money when disaster happens, you need to put more effort and money for disaster recovery action.
Regards to back up strategy, you need to make sure that your data is being stored somewhere on a regular basis. Let’s suppose you are a blogger and the web site stores your contents, images. On the other hand, your web site is interacting with visitors via comments and forms which need to be stored as well. In this case, you need to define basic rules for backup strategy;
- What kind of information will be backed up
- Which service/tool will be used to store data
- How often your data will be backed up
- How you make sure the data is backed up with success
Once you answer these questions, it is time to take action! Let’s suppose your data is being stored in EBS(Elastic Block Storage). For the backup strategy, one of the thing you can to do is to define Cloudwatch events in order to take a snapshot
Pilot light Strategy
Regards to Pilot light strategy, you need to keep running core components when disaster happens. When you take a look at the left side on the picture, core components should be tagged to prepare a pilot light strategy. In this case, the database server is selected as a core component. What you need to do is to create the core components in the AWS side as well.
In the Pilot light strategy, these are the main things that you need to consider;
- Same components should be created in the AWS side as well
- The components can be selected from a small instance in order to reduce the cost
- Core components need to be kept running. On the other hand, other components like a web server, the application server can be shut down. The point is that when a disaster happens, these components should be ready to start
- Data should be replicated from left side to right side on a regular basis.
When you compare backup strategy and pilot light, pilot light is more expensive than the backup strategy but it is a faster recovery strategy when disaster happens
Warm Standby Strategy
In this strategy, all components on the left side should be created on the AWS side. Regards to the ‘pilot light’, only core components are created on the AWS side, but the ‘warm standby’ strategy is asking to create all components on the AWS side. In this case, a smaller instance can be created on the AWS side, that means, when any disaster happens, the infrastructure is ready to use.
Scaling up strategy. All the components on the AWS side are created from the small instance, in case the traffic is rotated to the AWS side, the infrastructure is supposed to have a scale-up strategy.
From a budget perspective, ‘warm standby’ strategy is more expensive than the ‘pilot light’ strategy. The good thing is to note that your IT infrastructure is enabled in a quick time in the ‘warm standby’ strategy.
This is the most expensive solution when you compare with others. Every component on the left side should be created on the AWS side as well. This approach is very similar to ‘warm standby’ strategy, the main difference is that multi-site accept real traffic as well. For example, %80 of the traffic can go to on-prem and %20 of the traffic can go to the AWS side. It totally depends on your traffic division strategy. Another point is to note that, the multi-site strategy helps you to test your disaster recovery plan for 24/7
When to select the type of disaster recovery strategy, you need to make sure how your business continuity is crucial. Let’s suppose you have an e-commerce site and disaster happened. Even if downtime takes seconds, it means that the loss is happening for your business and profit.