Disaster Recovery (DR) is a simple concept that unfortunately gets quite complex real quick. At a high level, disaster recovery ensures the persistence of critical aspects of your business during or following a disaster, whether natural or man-made. Generally speaking, disaster recovery is focused around Information Technology, with a more holistic organization plan covered by a Business Continuity Plan.
The concept is simple enough. But how does one achieve this persistence? That’s where things can become very complex. But fear not, dear reader. Anexinet is here to untangle things. As part of our continuing DR blog series, two posts will focus on some core components of DR: Infrastructure and Policy/Procedures. This blog focuses on Infrastructure. Our next post will focus on DR policy and procedures.
With regard to DR Infrastructure, when most people talk DR they want to get right into the specific nitty-gritty: what are the most optimal EMC Recoverpoint parameters? What’s the most ideal configuration for Oracle Data Guard? And that’s all well and good, but let’s worry about the cake first, and the frosting later.
So, let’s take it up a few levels. Within Infrastructure, you have the Systems (prod & dr), the connectivity, and the data (hot, backup, replicate/off site).
Three major components of DR within Infrastructure
These include servers, compute power, and non-human workhorses. You use these to process your orders, make decisions, and process your business-critical data. They may be physical, virtual, or in the cloud, but you know each one by name. You start your day by logging into one and every step of your workday involves some server doing something to assist you. Without it, you lose your customer-facing website, you lose your applications, and you lose everything else that makes your business an efficient organization.
Disaster Recovery Systems
If your production systems had a twin, the DR Systems would be it. These are duplicate, regularly tested, and fully capable systems, able to take over all the work you depend on your production systems for, the moment a failure occurs. Ideally, your DR Systems are housed in a different facility than the production system and are able to run at full capacity with no assistance from the production systems.
This is how everything talks to one another. Your production systems are connected by at least two separate network switches. If you use a SAN, you will have two separate fabrics. If you use the cloud, your connection to the cloud will also be redundant. Any secondary offices, remote data centers, or remote locations also use redundant network connections. Any replication flows over these lines. Your network provides connectivity to all your production and DR systems, such that end users can access their data, systems, and applications seamlessly, regardless of the state of your environment.
The Hot Copy
This is the data your business depends on the active dataset that your applications, users, and databases read and write-to each day. Typically, this data is raid-protected, but further protections are necessary to ensure the data is safe.
The Backup Copy
This data set can exist in many forms, including backup storage array, replicated storage, checkpoints, journaled file systems, etc. It is meant as a low Recovery Point Objective option you can quickly use to restore data to handle non-catastrophic recoveries.
The Offsite Copy
This data is for long-term storage and is usually kept on a different medium than the Hot Copy and Backup Copy, including on tape, on removable media, in the cloud, or on a dedicated backup array. This data should be stored offsite and tested regularly. Additionally, this copy should be able to restore the data independent of any existing infrastructure and can be used to recover from a full disaster.
With those three areas identified, your business may begin holding strategic planning sessions to determine exactly which technologies and DR path are most appropriate for your organization and applications. Anexinet would love to help make your journey to a robust disaster recovery plan as smooth and successful as possible. To that end, check out our Disaster Recovery Kickstart: s strategy and roadmap that enables you to upgrade your DR plan and eliminate vulnerabilities. And look out for the next part in our blog series, where we discuss policies and procedures—the less sexy, but perhaps more critical aspect of a successful DR.