Cloud Hosting

Some people are born entrepreneurs. Nothing makes them happier than creating something new and innovative a product or service that challenges convention and disrupts traditional markets. They love the rollout, the market splash and reveling in the disruption. The app economy has brought entrepreneurship to new heights and along with it the accelerated adoption of cloud computing and a sea change in the foundational elements of IT. Once the initial heat from a new apps development and launch phases is over, when an app or service is on the street, its support becomes a matter of day-to-day business operations.

In the argot of industry, applications are typically developed and brought to market in three phases build, deploy and run that can be equated to the phases of the software development life cycle: day zero, day one and day two. They are not, of course, calendar days. A day zero effort could last for months, while a day one effort could take weeks or longer. But day two could last for decades however long the product, service or application remains on the market. Day two operations describe the monitoring, data management, maintenance and troubleshooting that keep a businesss apps, services and hosts up and running.

Of course, capably running day-two operations is fundamental to the success of any business. Its what turns entrepreneurs' great ideas into ongoing streams of customer satisfaction, company revenues and investor dividends. But its not always very creative. In fact, day two success often hinges on adhering closely to a set of inflexible procedures. Thats not satisfying for everyone. Take, for example, Jeff Bezos, for whom day two actually represents failure.

In his annual letter to shareholders, Bezos posed this rhetorical question: What are the techniques and tactics to avoid becoming a Day 2 company, and how does a large company maintain the values of Day 1?

"I dont know the whole answer, was his response. But I may know bits of it. He then went on to explain four different strategies, giving examples for each from inside Amazon.

I work in an Amazon building named Day 1, and when I moved buildings, I took the name with me, he said. Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1 [at Amazon].

However, theres more to day two than routine drudgery. There is also an assortment of key technical challenges that play a critical role in day two operations issues like ensuring uptime, continuity, resiliency, performance, compliance, scalability and security. These issues can sink a business if theyre not handled appropriately. In most organizations, the DevOps, CloudOps and IT security teams are the unsung heroes keeping the organization away from negative headlines.

Perhaps the greatest day-two challenge is dealing with or preparing for an outage or a full-service disruption. For many cloud-native innovators, theres a temptation to believe that one couldnt happen because the cloud is so resilient.

Thats only true up to a point. Any company that builds its applications in the cloud using technologies like containers and Kubernetes to take full advantage of the agility, performance and scalability of cloud-native strategies which is now the prevailing pattern among application developers should know that the cloud providers responsibility for security ends where the clients ecosystem begins. Its the clients business, not the providers, thats responsible for the data and processes happening inside those containers.

There is another commonly held misconception that data replication provided by underlying cloud constructs should be sufficient to avoid and recover from service outages. However, replication does not help against data corruption or loss. Data loss might be accidental or malicious especially in todays environment of increasing cyberattacks and ransomware. It is again up to the applications and the day-two teams to ensure that the wrong or corrupted data is not mindlessly replicated by the underlying cloud constructs.

Whatever its cause, should an application or service be distributed, getting it back online (or getting a business operational) is an urgent day-two challenge in need of a solution that creates continuity. Unfortunately, legacy backup and disaster recovery tools that have been used for years in old-school data centers for old-school operations dont work in cloud-native environments.

Containers and Kubernetes, which power a lot of the cloud-native environments, require that backup and recovery tools treat the application as the unit of atomicity. But thats only halfway to addressing the day two challenge. The other side is finding tools that make it extremely simple to operate at scale with security and self-service portals so that our unsung day two heroes can also have a good nights sleep!

Go here to see the original:
You're Already On A Cloud -- What Could Possibly Go Wrong? - Forbes