In Part I, we started with an overview of disaster recovery cloud delivery options and some of the more popular use cases.  In Part II, we will provide more depth on available disaster recovery cloud options and determining the right solution.

A good place to start is to highlight what events are most likely to create a disaster recovery scenario.  We often talk about disaster recovery is only in the event of a catastrophic event but as you can see from the table below most causes are more mundane. 1

Most of the disaster recovery events we have experienced have been facilities related issues such as generator failure, water damage or some other building related issue.  Cyber Crime is a significant issue and was the 2016 largest increase in root cause outage with a trend we expect to continue over the next few years.

Disaster Recovery Terms

Before we look at different cloud delivery options you may find it would helpful understand some commonly used terms and their definitions.

  • DRaaS (Disaster Recovery as a Service) – is a service offering that provides infrastructure and connectivity to facilitate the replication of workloads and recovery of workloads at a cloud provider. The cloud provider also assumes the environmental, logical, physical and financial responsibility of the target solution.
  • RTO (Recovery Time Objective) – describes how long it will take to get an application online.
  • RPO (Recovery Point Objective) – describes the shortest point in history the application goes back to when it comes back on line.
  • SLA (Service Level Agreement) – a contract between the customer, end user, department, or organization and a service provider (could be internal technical staff or an outsourced entity) describing the availability of an application, infrastructure, or enterprise.
  • SLO (Service Level Objective) – SLOs usually exist between the collective RTO and RPO time, and that of the SLA. An SLO is an objective and is usually more aggressive than a contract or an SLA.

Leveraging Public Cloud for Disaster Recovery

Gartner defines public cloud computing as a style of computing where scalable and elastic IT-enabled capabilities are provided as a service to external customers using Internet technologies—i.e., public cloud computing uses cloud computing technologies to support customers that are external to the provider’s organization.2

Public cloud disaster recovery provides replication services as either a replication target from your source equipment at your location or a replication target from your workload already in the cloud.  There can be financial and operational benefits to leveraging Public Cloud for disaster recovery.  The financial benefits include leveraging the massive economies of scale that the large Public Cloud providers can offer.  Efficiencies and increased competition will continue to bring downward pressures on the price of Public Cloud.  You can also realize significant operational savings by shifting the administration of a secondary site to a cloud provider.

There are benefits in leveraging a Public Cloud but also some potential drawbacks you need to take into consideration.  One drawback to consider is that you lose some of your operational identity when you go to a Public Cloud provider.  The way they achieve economies of scale is to have all users (tenants) of their cloud conform to their process and procedures.  If you have tailored your operational process to the needs of your specific business, you will find they will be compromised to fit the Public Cloud provider’s ways of doing things.

A second drawback is the loss of technical flexibility; you can replicate to their stuff as long as you ensure your replication technology is something the provider can support.  For example if you have a large investment in a certain storage technology and part of your replication strategy is leveraging storage replication you need to understand what your Public Cloud provider replication target options are.  A third drawback is that with some Public Cloud providers the cost of replication is inexpensive it is the cost of recovery or the cost of failing back where cost can become significant.

Pro:  Inexpensive replication target, operational savings.

Con:  Lost operational flexibility, hidden cost of recovery.

Bottom-line:  If you can conform to the operational and technical demands then you can save a significant amount of money.  If you have, a very diverse datacenter or the business cannot support significant operational or technical changes you should consider another option for disaster recovery.

Private Cloud for Disaster Recovery

Private cloud computing is a form of cloud computing that is used by only one organization, or that ensures that an organization is completely isolated from others.3  Clients either look to deploy private clouds in a client owned secondary location or in a colocation facility.  With the amount of data center space available we typically run into firms that want to leverage a collocation facility.

Co-location, by definition, is multiple entities functioning from within the same physical location. Co-location facilities typically provide environmentals for space, power, cooling, and physical security for the server, storage, and networking equipment owned and operated by their clients. Since the overall physical data, center cost is shared with all the co-location tenants, operators can offer this type of service to clients for much less than clients can build or secure their own facility.  We see clients leverage co-location if they do not have a secondary site of the secondary site is too close to survive a regional disaster.

Pro:  Greater control, more flexibility.

Con:  Greater expense, greater operational responsibility.

Bottom-line:  If you have the physical location and datacenter environmentals at a secondary location that is regionally dispersed this could be a preferred option.  A private cloud provides the opportunity to deploy your disaster recovery solution closer to the way you would like without much compromise.  If you do not have a second location, the locations are too close or you have budget limitations this could be a challenging option.

What is “Custom Cloud” and how to leverage for your Disaster Recovery

A custom cloud deployment tries to leverage the best that public and private cloud have to offer.  The fundamentals of a custom cloud is the ability to co-exist shared resource pools with dedicated assets for a total solution.  The shared pool would be assets that the cloud provider has that are consistent with the client’s current environment.  The dedicated assets are ones that the cloud provider makes available to the client but does not have as a part of a shared service.

The key to a successful custom cloud deployment is the technical and operational integration of the shared and dedicated resources.  Part of the engagement with your custom cloud provider is to map your solution set with their shared service offering.  After the mapping, you should have a clear picture of which cloud resources you can use and which resources you cannot not use and what the gap is for a total solution.

The next step would be to determine if the gap can be solved by conforming to the solutions providers’ asset pool or if a dedicated solution is needed.  If a dedicated solution is needed then a detailed physical and logical interoperability analysis is completed.  If the interoperability analysis comes back clean, then discussions can begin on the configuration requirements.  Once a configuration is agreed upon then the physical planning process to co-locate the dedicated asset in the cloud provider’s datacenter is completed.

The other critical aspect of the custom cloud solution is the operational integration.  From the mapping analysis, you may have dedicated asset requirements and if that is that case the next conversation is who should manage those dedicated assets.  The cloud provider may have the skills to manage dedicated assets but sometimes they do not.  If the cloud provider does not have the skill then the client will have the management responsibility of the dedicated asset.  Split administration of the target assets can and does work the key is in the trust you have in your provider.  If you clearly define roles and responsibilities, work collaboratively, communicate often and trust your provider this will be a successful deployment.

Pro:  Less Expensive than Private Cloud, more flexibility than Public Cloud

Con:  More Expensive than Public Cloud, Less flexibility than Private Cloud.

Bottom-line:  If you are unable to deploy a private cloud and find public cloud too restrictive then this is a viable solution.  Find a provider you can trust since this will often be a technically and operationally collaborative solution.  You can realize the operational, technical and financial benefits of cloud with minimal compromise of your technical and operational knowledge base.

Caution

With all of these options available, a troubling trend we are seeing is that clients are choosing them all!  A little public cloud for this, a little private cloud for that and a little custom cloud for all the weird stuff in my data center.   There are many problems with this approach including increased complexity and unaccounted application interdependencies.  The complexities of coordinating a multi-system failover across multiple DR providers significantly reduces failover and failback success.  The other issue of a multi-cloud DR approach is with application interdependencies.  If your systems have been around for a while, you have code connecting one application with another and if these systems are protected differently then you will not be successful with your application RTO.

Conclusion

Cloud utilization is pervasive in the market the danger, for disaster recovery, is choosing the wrong deployment method and negatively affecting your business.  Understand your systems, IT talent, processes and external influences to determine the right cloud solution.  Public cloud is a strong financial consideration if you can conform to their solution stack.  If you cannot fully compromise your current IT environment to that of a public cloud provider then the solutions come down to a private cloud or a custom cloud.  Security, personnel and financial impact are the major criteria for one versus the other.  If you look for help with an overall DRaaS analysis, make sure to leverage a partner that understands and has experience with public, private and hybrid cloud solutions.  We will talk a little more on partner decision criteria on the next blog.

Research

  1. Whitepaper available through Zerto; Cost of Data Center Outages, January 2016; Data Center Performance Benchmark Series
  2. http://www.gartner.com/it-glossary/public-cloud-computing/
  3. http://www.gartner.com/it-glossary/private-cloud-computing/