fbpx

What is Cloud Computing? Part 5 – Cloud Challenges

In the first 4 parts of this series, we defined the Cloud and discussed some of its many advantages. Which brings up an interesting question—“Are there any potential pitfalls of migrating your business to a Cloud Computing environment?” We’ll explore that issue in this post.

One issue that has to be addressed when planning a Cloud migration is the connection to the Internet. Most businesses only have a single Internet circuit. If the Internet circuit goes out, most employees can still do a limited amount of work on their personal computers (PCs). As long as the Internet outage doesn’t last too long, it isn’t a debilitating problem (some would argue that due to the proliferation of cloud-based applications like email and Google Docs, this statement is debatable).

But in a Cloud Computing environment based on desktop virtualization, an Internet outage means that employees do not have access to their desktops until Internet service is restored. Obviously, that is not a good situation. So how can this risk be mitigated?

The answer is dual Internet connections for your business. Today, most small and medium sized businesses (SMBs) in major metropolitan areas can order a second 20 Mbps Internet circuit for less than $100 per month. All business-class Internet routers can be configured to use dual Internet connections (often referred to as “dual-WAN” by the propeller heads) with fail-over to a backup connection if the primary connection fails.

In the near future, businesses may be able to use the 5G network as a secondary Internet connection. Routers will likely be designed with built-in 5G modem slots or external modems can be used to convert the 5G signal to Ethernet. Either way, the flexibility of the 5G network can be brought to bear on this issue. The following article gives a good overview of the current state of 5G deployment in the US and its potential capabilities.

The other potential problem is the Cloud itself—what happens if the Cloud goes down? Is that even possible? The following article lists the 10 biggest “Cloud” outages of 2019 (so far).

Let’s break these outages down:

  • 7 of the 10 outages were Cloud-based software services (software as a service—SaaS). The majority of these issues were caused by routine maintenance procedures, software system updates or 3rd-party support issues. One could argue that these are not “truly” Cloud outages, but rather software service outages.
  • Of the 3 remaining issues, 1 affected Google, 1 affected Azure and the 3rd affected AWS. For Google and Azure, core Cloud services were affected. For AWS, there was a bandwidth issue related to Internet Service Provider Verizon that did not directly impact core AWS services.
  • The majority of these outages lasted less than 4 hours.

It is worth noting that the issue experienced by AWS on June 24, 2019 did not affect any of Red One NS’ clients. That is due to the fact that AWS has data centers in multiple regions throughout the world and none of our client systems were in the affected region. Perhaps even more important, the outage was not caused by AWS and once the Internet routing issue was resolved, all AWS services were readily available.

To be sure, AWS is not immune to network issues. However, as Mai-Lan Tomsen Bukovec, VP and general manager of Amazon S3 noted at last year’s AWS re:Invent conference in Las Vegas, Amazon Web Services (AWS) has never seen an entire data center fail within its massive global infrastructure footprint.

This is a fact that Red One NS relies on to ensure that our clients never experience a prolonged Cloud outage. In the event that an AWS data center hosting one or more of our clients’ servers goes offline, we can rebuild those servers in another data center in the same region or even in a different region, if necessary.

As previously noted, most outages are caused by updates and routine maintenance which could just as easily occur in an in-house data center. In fact, because Cloud providers have more network resources than the typical business, the duration of Cloud outages may be significantly less than similar outages at Fortune 500 companies not using the Cloud.

So, is Cloud reliability an issue or not? I suppose that depends on how you define Cloud computing. If you use the Cloud for SaaS, then you are certainly reliant on the service provider to make sure their platform is available on a consistent basis. Given that this is the segment of Cloud services where most of the “outages” seem to occur, it is definitely an important consideration. On the other hand, if you have been following this blog series, you might agree that this really isn’t a “Cloud” problem.

If you are using the Cloud to host all of your organization’s desktop/server compute and storage, then Cloud availability is certainly a critical issue. But with proper planning and testing, the frequency and duration of any type of IT outage can be minimized to an acceptable level and Cloud Computing is not an exception to that rule. The key to having a successful Cloud-based IT infrastructure is having a team of Cloud professionals who understand the risks and have the knowledge and experience to mitigate those risks to an acceptable level. From that perspective, the Cloud isn’t that much of a paradigm shift and the small risk of Cloud reliability shouldn’t prevent any organization from reaping the tremendous benefits that it offers.

In the next installment of this series, we’ll look at the tremendous flexibility that Cloud Computing offers via its Scale on Demand architecture.

Leave a Reply

Your email address will not be published. Required fields are marked *