Electronic Health Reporter

Guest post by Shannon Snowden, senior technical marketing architect, Zerto.  

Electronic health records (EHR) are the tie that binds together the patient with the caregivers. What happens when an extended outage or disaster happens? Caregivers still have to administer treatments regardless if the systems are online.

The longer the outage, the greater the negative impact to the quality of the end product or service. In the healthcare business, it is unacceptable. Every manually tracked record has to be added back into the EHR when it is available once again.

A big concern is that the manual records often get summarized with many of the details those electronic healthcare systems track are missing. These knowledge gaps ultimately could diminish the quality of patient care.

A contributing factor to the difficulty in finding a good disaster recovery solution is the technology necessary to support healthcare information systems (HIS) are complex, involve multiple servers that are tightly integrated and are quite unique from the perspective that the application vendor remains very involved with the customer on an ongoing basis.

This is the challenge faced by healthcare organization CIO/CTOs, IT directors and managers. How do you provide a sound business continuity solution that enables nearly no interruption in patient services is easy to manage and is within a realistic budget? What should be considered requirements for a healthcare information system disaster recovery solution?

Here is what to look for in a disaster recovery solution:

It must be hardware agnostic. The disaster recovery solution needs to be able to accommodate a variety of hardware. It is more likely that different data centers will have completely different hardware, so having a disaster recovery solution that supports any hardware makes the data protection solution very flexible.

Continuous Data Protection (CDP) replication. CDP is a radically different from traditional solutions for disaster recovery. The traditional solutions were either simply backing up the data to tapes or using the storage hardware to take a snapshot of the entire storage volume and then replicate that to another SAN of exact hardware. CDP replicates the data as it is being written to the storage at the production site to the recovery site as it happens. This ensures no information gaps in the patient record.

A DVR for the datacenter. The CDP solution should have the point in time (PIT) journal that effectively acts like a DVR for the datacenter. PIT allows for the recovery of a single server or the entire datacenter and supports the recovery going back to any time within the last several previous days.

Protecting the servers into logical affinity groupings. This allows them to be treated as a logical entity that supports an application or a service tier and can be failed over at the same time. Consistency is critical to ensure applications can be recovered which makes affinity groups a key part of any disaster recovery solution.

Don’t settle for a point solution. The same software driving your disaster recovery can also help you meet other requirements, such as meaningful use, by allowing you to migrate to new systems with just a few minutes of outage.

Meets Recovery Point Objectives (RPOs) of less than a minute and should really be able to support and RPO of just a few seconds.

Meets Recovery Time Objectives (RTOs) of less than 15 minutes for an application and an hour or less for an entire datacenter.

Non-disruptive testing. This is critical. If the data protection solution can’t be tested on a regular basis without impacting the production environment, it’s not keeping up with competitive products. Keep looking, because these products exist and are very affordable. Non-disruptive testing directly impacts RTOs. A good disaster recovery solution allows for the complete orchestration of a failover event to be regularly tested without impact to the production environment. Non-disruptive testing lets administrators perfect server boot order, boot timing of servers, IP address changes if needed, pre-or post-boot scripts be executed.

Easy and intuitive to install and use. It shouldn’t take a professional services engagement to design and deploy a disaster recovery solution. If it does, then you probably will experience difficulty in maintaining and upgrading the environment, which will significantly impact the productivity of IT staff. Modern data mobility solutions installs in less than an hour by administrators with basic skills.

Future proof. Avoiding infrastructure technology vendor lock-in, like a specific storage array or hypervisor, has many advantages, such as keeping you in charge of your own datacenter and the infrastructure vendors have to stay competitive. Being able to leverage private, hybrid or public cloud solutions with the disaster recovery solution only strengthens negotiation power with providers and maximizes the solution options, which drive down costs.

Once a modern data protection solution is in place, it will support a complete site failover with a push of a red button, even by someone who may or may not be a regular administrator. That’s because the non-disruptive testing allows for the failover workflow sequencing to be perfected during the middle of the day and can be done as often as needed with no production impact.

It can also become transformative and more than just a disaster recovery solution. It can be integral in supporting business activities, such as acquisitions and mergers.

Not having hardware dependencies allows for completely disparate IT systems to be connected and act as recovery sites for each other with bi-directional replication and orchestration. It also can streamline datacenter consolidation by allowing the data to be replicated with no impact to the production environment and then the migration can be test non-disruptively to ensure the datacenter move outage window is very short and successful.

Disaster recovery solutions historically have been a major constraint, inflexible and expensive. Recovery testing either didn’t happen or rarely happened to the point that nobody had confidence it would work and everyone hoped that a disaster just never happened.

That all has changed. You now have more flexible and capable choices than ever before and can transcend simple disaster recovery point solutions into a proactive business enablement tool.