Image: Burnt out server room

What happens if my applications aren’t available?

For most organisations, the delivery of IT services and applications is crucial to the running of the business. So disaster recovery (DR) is critical. Non-availability of these capabilities can affect sales, cause brand damage and legal issues, or even result in job losses. Clearly, then, it is important to manage what happens if components, servers, applications or even data centres fail.

Levels of protection

To protect your organisation against loss of service, there are three primary solutions that can be implemented. Backup, high availability and disaster recovery protection. Backup is considered the minimum level of protection. Your data is restored in the event that it is deleted, corrupted or otherwise becomes unavailable – important with threats such as ransomware. For applications critical to the business, such as e-commerce, finance, public web sites and ERP, high availability is essential. This is built into the service to ensure that it will continue to work in the event a component or server fails. Neither of these solutions, however, will provide protection against a more serious situation. Imagine a data centre failure when all services are lost, potentially for a long period of time!

The challenges of disaster recovery solutions

The impact of an outage can be very damaging, so you have to have a disaster recovery capability. However, until now it has meant additional costs in hardware and/or third party services, plus a significant burden on IT. They are often sent to a recovery location for a week with hundreds of tapes to recover the services as best they can. Typically test results are partial recovery and/or services not being tested fully because it is too difficult. In the event of a DR invocation, it may take a long time to get services up and running and some may not be available. Which is why some organisations take a best endeavours approach if a failure situation occurs. However, they may consider this an acceptable risk, auditors, insurance companies and compliance officers may take a different view.

How can Microsoft solutions help?

Image: Azure Site Recovery Logo

The Microsoft Azure cloud platform provides many services, functions and capabilities. This includes deployment of ‘infrastructure-as-a-service’ virtual machines similar to on-premises solutions such as Hyper-V or VMware. In Azure, it is possible to build hundreds of servers from a script in a few hours. Entire data centre deployments can be stood up in a few days (with relevant designs and preparation in place). However, this doesn’t solve the issues of DR protection. This is where Azure Site Recovery (ASR) comes in to play. At a basic level, ASR creates replica copies of servers enabling them to be turned on in the event of a serious outage.

What are the benefits of these services?

The key benefits of Azure include reduced costs and complexity with improved flexibility & agility. There are no long procurement cycles, no need to purchase any additional hardware or secondary sites and you only pay for what you use. The Azure platform has been designed to be resilient, reliable and secure, typically more so than even the best configured on-premises data centres. From a security and compliance point of view, the key certifications are in place, including ISO 27001 for security operations, 27018 for personally identifiable information and PCI DSS. So you don’t have to worry about being up-to-date. ASR itself is delivered as a service, requiring no maintenance, patching or upgrading. Together, Azure infrastructure services and ASR enable a level of DR protection and flexibility that is not typically available using traditional or legacy solutions.

ASR in action

Image: Site Recovery options covered by ASR

ASR’s primary function is to replicate servers that could be running on Hyper-V, VMware, physical machines or other cloud platforms. Supported guest operating systems naturally includes Windows but also a large array of Linux distributions from Red Hat to Ubuntu. Protected machines sit dormant in a recovery vault until they are needed to recover from a disaster situation.
When DR is invoked, ASR orchestrates and manages the failover to re-hydrate the machines in the correct order, re-addressing systems where necessary and allowing for any manual steps to be completed as part of the process. When the DR situation has been rectified, ASR can fail back the systems to their original locations.

Does ASR do anything else?

Aside from delivering its primary DR function, ASR also supports server migration into Azure and provides the ability to easily deploy dev/test machines as exact copies of the production server into an isolated network. Whilst not as flexible as a full DevOps-style dev/test/production solution, it may suit your requirements and is significantly easier and cheaper to implement.

How is ASR Licensed?

Each machine that is replicated by ASR requires a licence. In the normal running state, this and the storage for the images are the only costs.
When a DR failover is initiated and the machines are re-hydrated, the normal machine running costs will be applicable.

Are there any other considerations?

A supporting requirement is for Microsoft applications to have Software Assurance attached. This has a related Disaster Recovery Right, which allows these applications to be replicated into Azure, enables testing and allows failback within a 90 day period. Without this right, at least from a licensing point of view, it is not possible to test disaster recovery or failback until after 90 days, two aspects that are particularly important in the context of disaster recovery. It is also important to note that the Disaster Recovery Right is only valid when using Azure Site Recovery. This is  indicated below by this extract from the Microsoft Product Use Rights document:

Image: Disaster Recovery Rights Terms

This implication of this is that the use of similar, non-Microsoft replication solutions will result in the loss of essential DR capabilities.

Table: Features with and without ASR

Webinar

On 6th February 2018 Silversands Consultants Gary Jensen and Michael Coutanche will be hosting a 30 minute webinar on this subject. In the webinar you will have the chance to see a quick demo and ask questions. Also, as an attendee your business may qualify for a free half day of consultancy. Register here

How can Silversands help?

Having worked with Azure since the early days of this cloud platform, Silversands has built up significant experience and knowledge. As such, we are ideally placed to assist you with an implementation of ASR to support your disaster recovery, migration or dev/test requirements. In addition, there are countless other Azure-based solutions ranging from infrastructure services to platform and custom applications we can support. To start a conversation about Disaster Recovery solutions for your organisation please complete the form below. Also we’re very active bloggers and commentators around Microsoft Azure, Office 365 and SharePoint, so please do follow us on Twitter and LinkedIn.

Icon: LinkedInIcon: twitter bird

Contact us

  • This field is for validation purposes and should be left unchanged.