For those outside the world of disaster recovery and business continuity, the following passage from the recent Uptime Institute report, “Post Pandemic Data Centers,” may be a little unsettling:
"There is an important assumption running through our analysis and, we believe, in the planning of most operations: COVID-19 will almost certainly not be the last pandemic — and it may only be one of many."
Most people have been blissfully unaware of the impact a global pandemic could have until this year. Now, as they battle impatience at having to continue to deal with the restrictions required to control the virus, they may have to face the reality that this might not be a “once in a lifetime” event. We could be living under the shadow of a pandemic from now on.
As the business continuity manager for an essential supplier of critical infrastructure and services to the data center industry, I don’t have the luxury of blissful ignorance. Our business continuity depends on the continuity of our data centers, including a network of cloud and colocation providers. And our continuity enables the continuity of our customers’ data centers. Our plans and activities have to account for the critical role we play in the industry and the critical role the industry plays in continuity, especially during a pandemic.
The new normal for business continuity
Pandemics not only create distinct challenges to business continuity; they increase the demand for data center applications as everything from shopping to weekly staff meetings moves from the physical to the virtual world. They also force organizations to operate within the response stage of their disaster recovery plans for months or years rather than days or weeks. And they don’t occur in a vacuum. This year alone Vertiv, which has operations globally, has dealt with more than 20 disaster events in addition to the pandemic.
That means for most of this year we’ve been in dual-disaster mode as we deal with short-term events such as hurricanes and wildfires while simultaneously maintaining our response to the pandemic. This could represent the new normal for business continuity.
Best practices for continuity
Maintaining data center continuity during a time of dual disasters requires collaboration across data center-dependent organizations, their cloud and colocation providers, and data center equipment and service providers. Following are five of the best practices that have enabled us to maintain continuity for ourselves and our customers during this time of dual disasters:
The pandemic was on the radar of business continuity professionals beginning in December 2019. As part of our planning process, we worked proactively with our cloud and colocation providers to ensure preparedness. For example, we planned to implement a virtual desktop infrastructure in advance of the pandemic to facilitate work from home and we needed to be sure our partners could provide the capacity necessary to support that initiative. We communicated with our customers as early in the planning process as possible and tried to anticipate their needs. We expect our computing partners do the same with us.
Fortify power redundancy
The first thing I look at when evaluating potential data center partners is the power architecture. Do they have the necessary redundancy, backup power systems, and processes in place to provide power continuity in event of grid failure? One thing we don’t need today are any additional continuity challenges from power outages caused by preventable issues such as uninterruptible power supply (UPS) battery failures or a generator that fails to start as planned. We also want to ensure we have distributed our resources geographically to minimize our exposure to a widespread power grid failure.
Continued monitoring of our network has been instrumental in our ability to respond to the capacity fluctuations that have occurred during the pandemic. By always knowing where we are in terms of capacity, we’ve been able to work with our providers to spin up new capacity in a timely manner. We expect our data center partners to be doing the same with their critical infrastructure. The increased implementation of equipment and facility monitoring and the move to “smart hands” management have proven valuable to data center operators and their customers during the pandemic.
The availability of cloud services has been a lifesaver for us during the pandemic. But relying on cloud services means we have to compete with other organizations for available resources as our need for capacity grows, much like it did during the pandemic. We managed that challenge by prioritizing our needs and working closely with our partners. We also had to be sure we could provide the industry with the critical infrastructure required to continue to expand available capacity. That includes our ability to manufacture pre-engineered data center modules and facilities in a controlled factory environment. These prefabricated systems allow cloud and colocation providers to bring new capacity online faster than using traditional processes while reducing the complexity and labor required for deployment.
Plan for service
Specialized services, such as those required to maintain data center power and thermal infrastructure, can be vital in the safe recovery of data centers affected by natural disasters. But when you’re dealing with a pandemic plus a natural disaster, it may not be possible for specialists to travel to the affected area. An important part of our business continuity plan is ensuring we have service specialists available locally to assist our customers in the recovery process.
Business continuity doesn’t happen without data center continuity. In my role at Vertiv, I’m responsible for ensuring we have the data center resources we need to maintain continuity whether we are dealing with one disaster or two. However, that goal only matters in the context of a larger goal: ensuring our customers have access to our engineering, manufacturing, and service resources when they need them most.
More from Vertiv
Conference Session How to power 10MWs at the edge: What have we learnt so far?