With locations smack-dab in the middle of the San Andreas Fault, Silicon Valley, and Bay Area, enterprises and the data centers which serve them must pay special attention to disaster recovery and business continuity processes. In fact, many of these processes have been embedded since the costly devastation of the 1989 California earthquake.
The rudimentary guidance to area businesses came from California and FEMA immediately after the earthquake:
- Develop a business continuity plan
- Conduct an audit of general earthquake vulnerability and a hazards risk assessment
- Establish a contingency plan
- Conduct a non-structural assessment of your business, including inventory
- Hold regular drop, cover, and hold on drills for employee safety
- Encourage employees to have family plans and emergency kits
- Seismically retrofit buildings or occupy/rent buildings built to earthquake code
- When considering a site for a new business, evaluate the site for hazards, such as risk of liquefaction and proximity to faults; locate near lifelines, such as transportation, power, and water; and build to earthquake codes.
This is good guidance overall. Many of those guidelines are still in place.
However, the nature of enterprises in California has changed so dramatically since 1989, that disaster recovery and business continuity efforts have evolved from drop-and-cover drills to “sustaining mission-critical businesses which keep the region open and running.”
Small businesses, enterprises, and hyperscalers in California have developed disaster recovery solutions from within at every stage of the IT process. Many are also demanding the same from their data centers.
What is certain is that enterprises need to review and battle-test their disaster recovery plans regularly — at least once a quarter. Data centers play an instrumental role in this process by offering data backup and offering other disaster mitigation measures which enable companies to ensure uptime reliability and data integrity.
Business continuity focuses on comprehensive planning for the organization as a whole to mitigate risk. Disaster recovery is typically part of the business continuity equation – but centered on IT infrastructure. Building and complying with a disaster recovery plan is key; it outlines the steps a business needs to restore technology systems after a disaster or an emergency event has occurred.
Gartner defines disaster recovery as (1) the use of alternative network circuits to re-establish communications channels in the event the primary channels are disconnected or malfunctioning, and (2) the methods and procedures for returning a data center to full operation after a catastrophic interruption (e.g., including recovery of lost data). The best enterprises and data center managers look at both definitions equally under the twin banners of business continuity and disaster recovery.
Geography is Key
To address geographic challenges, smarter enterprises in California are considering build-to-suit options for their data storage. One size does not fit all. In these cases, location and proximity are critical. Having a data center 50 miles from your primary data center as your backup is a great strategy unless it’s on the same fault line. Key among the questions for the data center manager: “Can my employees get to that site?”
It’s not just the proximity of core data — it's also the stability of the environment and the availability of power and connectivity. The closer, the better, but be far enough away to protect yourself.
In California, for example, it is important data is stored in a seismically stable area – preferably above the 500-year floodplain – but still close enough to major urban areas like San Francisco or the Silicon Valley to maintain low latency. In addition, a sustainable power source nearby is critical.
These questions of geography are existential questions, all of them incredibly complex — and for which there are easy answers. Every enterprise struggles with them. There is no one-size-fits-all. In fact, it’s most likely a combination of all of those examples in the enterprise. An enterprise must look at all of its options when considering a data center — everything from colocation, joint ventures, powered shell, build-to-suit, and sale/lease-back.
The enterprise must also look at wholesale data center real estate options that provide “partnership as a service” (PaaS) models to help organizations plan and prepare for disaster recovery.
A good data center team can then customize specific solutions within those options based on workloads and the nature of the data, whether it’s rich media, relational databases, or file systems – which all need to be treated differently.
Partnership as a Service as a Concept
In “partnership as a service” models, different from “platform as a service” models, data centers can construct a different approach to partnership by co-creating capital expenditure and real estate strategies with its customers that both advance their mutual physical business interest in the digital world and – just as important – increase per share valuations. These kinds of arrangements provide enterprises with ownership options and dynamic leasing models that define true corporate partnerships.
If you need a disaster recovery plan, you must test it. Don’t leave it on the desk collecting dust. This is where the notion of PaaS comes into play. Enterprise leaders responsible for IT disaster recovery should solidify the scope of their investment by evaluating requirements such as workload types and recovery location needs, as well as future strategic initiatives including application modernization, cloud migration, and automation initiatives.
Additionally, IT professionals should determine whether PaaS is more appealing than do-it-yourself approaches by deciding if time to value is a high priority and if they want their team to spend time setting up tooling, creating run books, and managing the target disaster recovery location.
PaaS, please meet DRaaS; DRaaS, PaaS
Disasters are not a question of if, but when, and the best partners in the PaaS landscape are also well equipped for disaster recovery. Those partners are often well known for traditional disaster recovery and have evolved service offerings. They distinguish themselves through decades of experience managing disaster declarations and proven competency in supporting the recovery of a wide range of workloads. They have the largest and most complex engagements. Those partners have combined global cloud infrastructure scale with replication and automation capabilities to provide DRaaS through a (software as a service) SaaS model.
It was predicted that disaster recovery budgets for enterprises would increase this year, with half of businesses expected to spend more than 7% of their IT budgets on disaster recovery. However, during financially challenging times due to the COVID-19 pandemic, disaster recovery is one of the first areas looked at for cost reduction.
The best prepared have industrialized their offerings and invested in value-added service components to ensure application recovery at all levels.
Gartner predicts 60% of infrastructure and operations leaders who have developed cost-effective traditional disaster recovery capabilities will expand their roles to address newly emerging areas of IT resilience by 2025. This is good news for disaster recovery overall.
Back to the Golden State
Thirty-two years after the big earthquake, the California Department of Technology updated its best practices for enterprises as they build their disaster recovery plans.
- Provide a highly available, reliable, operationally efficient infrastructure.
- Provide automated incident detection and service recovery from a single point of failure within five minutes at a single physical site.
- Ensure hardware and software redundancy.
This is now industry-wide guidance which doesn't stop at California’s borders or fault lines but informs every enterprise IT operation and data center in 2021 looking to guard its digital assets.
