Cloud infrastructure has become indispensable to digital-first businesses.
Gartner senior PR director Christy Pettey describes it as the foundation of successful digital transformations. In fact, it’s so important that many enterprises are now concentrating on building cloud-first strategies.
Netflix is a great example of a cloud-first organization, note McKinsey analysts Nagendra Bommadevara, Andrea Del Miglio and Steve Jansen: “[Netflix] spent seven years on its transformation, adopting a cloud-native approach, rebuilding all its technology, and restructuring the way it operated. It employed application program interfaces (APIs) to reduce its monolithic legacy applications into smaller components, make them more flexible, and then move them to AWS.”
The move has seen the company slash IT costs and creep closer to its goal of 99.99 percent uptime.
You don’t have to be a Silicon Valley unicorn to achieve almost-perfect uptime scores. Any organization can take steps to improve the resilience of its cloud systems and edge nearer to 100-percent uptime.
Step 1 is to decide on a cloud infrastructure model. From there, you can make tweaks and adjustments to optimize your systems.
Infrastructure Model Option No. 1: The Public-Private Hybrid Cloud Systems
If you are looking to build a resilient cloud infrastructure, hybrid clouds are certainly the more dependable choice, says Intrepid Net Computing consultant Brent Kirkpatrick. They are more reliable, have better uptime and even appear faster to end users.
A hybrid cloud infrastructure, explains Backblaze’s Roderick Bauer, is a “mash-up of on-premises and off-premises IT resources.” It is a cloud-based environment that combines on-premise private clouds with third-party public clouds. As a result, workloads and data can move freely between the two as demands change.
The design of your hybrid cloud infrastructure is essential, writes TECA Inc. President Michael Otey. If demand is unpredictable, then you need to allocate sufficient resources to handle peak loads. Many companies use the private section of their cloud infrastructure to run their operations and the public section to handle any surges in demand. You’ll also want to host applications across several cloud availability zones to avoid downtime as a result of a regional outage.
Hybrid cloud models are popular with some of the biggest companies in the world. Technical writer Lucie Lozinski explains that Uber uses a hybrid cloud model for 100 percent uptime. Using a mix of cloud providers and several data centers means that it doesn’t matter if one fails; another is always there to take its place.
Infrastructure Model Option No. 2: A Multi-Cloud Strategy With Public Cloud Systems
Even if you aren’t hosting your own private cloud, you can take steps to boost uptime. While public clouds are risky, using several of them at once can make your enterprise’s cloud infrastructure significantly more resilient.
Granted, no public cloud provider is perfect, writes ZDNet UK editor Charles McLellan. “All cloud providers — even hyperscale ones with multiple geographically dispersed, redundant data centres — suffer outages from time to time, so putting all your workload ‘eggs’ in one provider’s ‘basket’ runs the risk of a mission-critical application becoming unavailable,” McLellan writes. A multi-cloud strategy is the only way to guarantee resilience when you rely solely on public clouds.
“With a multi-cloud strategy, in the event that one cloud platform fails to respond or is unable to deliver optimal performance, applications can be run on an alternative cloud option,” Tableau Software SVP EMEA James Eiloart writes. This is similar to a hybrid approach, except you still rely on third-parties when one cloud infrastructure fails.
The key to a successful multi-cloud strategy is seamless integration between providers, says Gijsbert Janssen van Doorn, the director of technical marketing at IT resilience company Zerto. You want to be able to move or access data instantly when a problem arises or when downtime is scheduled with a particular provider. One way to do this, van Doorn says, is to use a multi-cloud resilience platform that allows engineers to manage upgrades and prevent downtime completely remotely.
3 Tactics to Increase Uptime
Regardless of what cloud infrastructure your enterprise has, there are a few things you can do further to increase cloud uptime.
Eliminate Single Points of Failure
All hardware is going to fail eventually, says the team at Total Uptime. You need to eliminate any potential impact on your infrastructure when it does.
In practice, this means two of everything — not just cloud providers, but routers, switches, power sources and servers, too. Whatever can go wrong, make sure there is something there to take over when it does.
Use DNS to Manage Workloads Across Providers
“DNS is one of the most powerful tools in the stack for managing workload,” says NS1 Founder and CEO Kris Beevers. “You can leverage the traffic management tools of modern DNS providers to weight traffic across cloud services, shift workload in response to real-time conditions and fail away from broken cloud service providers.”
You’ll also need to mitigate for DNS redundancy, too, Beevers writes. This will avoid downtime as a result of outages at major service providers. This can be achieved by deploying multiple DNS networks.
Do Your Due Diligence
Be thorough when researching your options for cloud providers, IT industry analyst Zeus Kerravala writes. Don’t buy for convenience, whatever you do. Kerravala warns that he knows several companies that “took the path of the least resistance” and regretted it later on.
Do your homework on any cloud provider you are considering. This means understanding historical and potential future performance, specifically in the area you need it. “The information on the vendor websites may not tell the full picture, so it’s important to do the necessary due diligence to ensure you understand what you’re buying before you buy it.”
Strive for 100-Percent Uptime in Your Cloud Systems
Outages are going to occur. Regardless of how resilient your cloud infrastructure is, downtime is almost certain to occur at some point.
Paessler’s Sascha Neumeier calls 100-percent uptime an “unattainable objective.” Even though services and data can be moved dynamically between different cloud infrastructures when systems are down or being updated, restarts are always required, and that leads to less than 100-percent uptime.
But even if 100-percent uptime is unattainable, there’s still every reason to try and maximize your cloud infrastructure’s uptime. Chief among them is the fact that downtime incurs huge costs.
According to research by the Ponemon Institute, the average cost of an unplanned outage is $9,000 per minute. According to those numbers, then, the difference between 99-percent uptime and 99.9-percent is more than $3.5 million.
What’s your cloud downtime going to cost you?
Images by: Cathy Yeulet/©123RF.com, John Schnobrich, Konstantin Pelikh/©123RF.com