The world is full of myths. From classic beliefs in monsters and unusual events to “folk wisdom” and “tried and true” advice that is based on nothing and often doesn’t help at all.
And the world of business technology has its own fair share of Loch Ness Monsters, Bigfoots, and “foolproof” plans for success. In this series of articles, I plan to bust a few of the more common myths that surround the cloud, building and optimizing applications for the hybrid edge of IT, and how performance issues in the cloud can impact not just websites, but any business-critical application or infrastructure.
So get ready for this episode of Cloud Performance Myth-Busters. Our myth for today is about enterprise applications, hybrid cloud infrastructures, and whether internet reliability issues have negligible or no impact on how these environments perform.
Is it true that, when building a data center and choosing a cloud provider and platform, the only performance concerns are uptime for the cloud provider’s data center and optimization for the on-premises systems? Let’s put that myth to the test.
If you’ve been involved in evaluating or choosing a cloud provider, they’ve probably talked about their performance in one way or another. They may have touted how long they’ve gone without an outage in their data center, bragged about how many points of presence they have around the world, and even touted the capabilities of the hardware in their data center.
This is all well and good— but the one thing they aren’t talking about is whether their cloud services were disrupted or even brought down entirely by some issue on the edge of the internet. Because the reality is that even if the cloud provider’s data center was running fine during the internet outage, yet for customers who couldn’t reach their applications and services at the cloud provider, the cloud provider was down.
The same is true for servers and applications running on-premises or integrated with the cloud and internet in a hybrid model. Powerful hardware, effective performance management, thorough code, and server optimization are all good ways to improve uptime and overall performance.
But again, the reality is that if something goes wrong on the internet, these great on-premises and hybrid systems won’t be able to prevent downtime or poor performance for customers and users accessing these services over the internet.
Of course, you might say to yourself, “How common are these big internet outages? At most I hear of a few each year, and they don’t affect everyone. What are the odds of this impacting my hybrid IT systems?”
Well, the odds are higher than you should be comfortable with. According to Aberdeen research into the impact of internet performance issues, 78% of organizations report that they have four or more disruptions to their internet presence each month, and 15% of organizations see 10 or more disruptions. If your business is like the majority, chances are you’re seeing more than a few downtimes each month.
Of course, you might next say, “But I’m sure most of these are short disruptions, and not some big hour-long outage.” Well, Aberdeen research also shows that 60% of organizations say that a minute of internet downtime costs them more than $1,000 (and 14% express the cost at more than $10,000 a minute). So, from this we can see that even short downtimes can add up quickly.
And then there’s the final myth: since your systems are primarily enterprise-focused and not really dependent on the internet, these outages don’t mean much. But the reality is that, unless your application is on a box that isn’t connected to anything else, odds are it is very much an internet-connected and hybrid application.
Businesses are astutely using the hybrid edge to provide lots of benefits, from more efficient disaster recovery to increased agility for development to improved access to capabilities not easily built or deployed in-house. And this means that even seemingly incongruous outages can impact APIs and integrations — and send vital services into a tailspin.
In fact, this occurred just this year when the Amazon S3 services went down. Since S3 is mainly about storage, many didn’t think it would have a far-reaching impact on applications and services. But when it went down, so too did many other applications and services that were relying on it under the hood of various applications and services.
In this modern world dependent on the hybrid edge of cloud and IT, our extent of interaction and integration means that everything is an internet service. And any outage or disruption on the internet can bring down even enterprise-focused services and applications.
And that means that the myth that the internet has limited to no impact on the performance of cloud and enterprise applications has been busted.
Stay tuned for our next article in this series, where we’ll take on the myth that there’s nothing businesses can do to solve performance issues and downtime caused by things happening out on the internet. We’ll find out if that myth stands up to reality, or if it too gets busted.