AWS primarily and Azure, of late, dominate today’s discussions around storage, backup and compute power. A quick glance at headlines from technology journalists, and a reader can glean a common coverage theme that ties these writers together — the ongoing discussion around the benefits of going all-in on the public cloud. However, in most cases technology journalists are writing about larger corporations, or big name installations, which may or may not reflect the actual trends taking place in the marketplace, especially at mid-size companies and organizations experiencing a growth spurt.
As one who is regularly engaged with the CIOs at mid-size and smaller companies and organizations, I don’t see them going all-in on the cloud right now; rather some are pulling back from it and either opting a hybrid cloud solution, or are going all-in with on-prem backup solutions. In fact, according to a survey published by SMB analyst firm Techaisle LLS, the hybrid cloud is now being used by 32 percent of midmarket (100 to 999 employees) organizations, and that figure is expected to remain relatively flat at 31 percent into next year, in spite of what your read in the press about AWS or Azure penetration.
In addition, hybrid cloud spending is predicted to grow 22.5 percent annually through 2021 to reach almost $92 billion. According to analyst firm IDC, spending on traditional on-premises IT infrastructure will remain larger than cloud spending over the next two years. Although, in aggregate, IDC expected that 32 percent of IT infrastructure budgets will go to external clouds and almost 11 percent to private cloud. The hybrid setup is currently the most preferred model among cloud users too; a popular 2016 survey from RightScale showed that 71 percent of cloud users operate a hybrid environment.
This trend has not gone unnoticed at Microsoft, which just announced the unveiling of its Azure Container Instances; a platform that makes it much easier for mid-size companies and organizations to get up and running with containers, effectively bridging the gap to a hybrid model that is both powerful, cost effective and easy to implement.
Here are the seven trendy reasons why I think mid-size companies and organizations are evaluating what is best for them and in some cases not going all-in public cloud services:
Several of the cloud service providers offer — out of the box — traditional intrusion detection systems (IDS) and intrusion prevention systems (IPS) infrastructure. These solutions primarily focus on preventing brute force physical entry or common network security exploits (e.g., guns, guards and firewalls). But when safeguarding valuable data or client information it is important to consider placing that data on-premise in order to gain greater control and oversight. As most CIOs know, it’s not a question of if their team will face a cyber security issue but when. It is always a good policy to have the ability to immediately get in touch with IT staff to detect, act and remediate an issue. No third party will ever care about another company’s assets and customers as much as it does. A third party might care about their own reputation and brand, but that doesn’t mean they will go above and beyond when you need them the most to protect your company, brand, and customer’s data.
Surprise cost overruns and budget variances are common in the public cloud simply because it’s hard to predict future application development mistakes. Developers are constantly under pressure to deliver new fixes and updates while still keeping the lights on. Invariably there will be a simple line of code that takes into account all of the caveats and benefits of running in a cloud environment while also having the ability to quickly scale, and as a result mid-size companies can suddenly be saddled with a large monthly bill that includes unexpected network surcharges, or unforeseen expenses for burst workloads or worse they forget to delete containers and virtual machines to scale down leading to severe under-utilization. With on-premise infrastructure, both the costs and asset depreciation are more predictable making budget planning easier.
Depending on where a company’s customers are located and the depth of its need for high performance and low latency access to data, a hybrid cloud approach is fairly commonplace. For example, web applications that support the company’s sales and marketing efforts may need to reside in the cloud, but an IoT device that measures another machine or onsite production processes can also generate vast amounts of raw data. That IoT device may need an accompanying onsite ingestion engine that can cache data locally before moving it to the cloud for long term archiving, or further analysis for business intelligence purposes.
I love the public cloud simply for its ability to quickly gain infrastructure resources on-demand. It’s very hard when a business suddenly grows 10x and needs to service customer orders rapidly, but then quickly falls behind the 8-ball in achieving those deliveries. That’s because company’s technology, processes, and workflows need to change and scale with the company. Doing it all “in the cloud” is fairly easy if all the business tools are already cloud-enabled (Salesforce for CRM, Quickbooks, etc.). There are less stakeholders involved in this model versus on-premise. However, if you have a custom legacy application that is not cloud ready it makes sense to keep it on premise since it may require proprietary hardware and ongoing OS upgrades and maintenance. The cost or risk of going all-in with the cloud exceeds the benefits of scale and agility for mid-size companies and organizations in growth mode.
Business, customers, and markets move at a rapid pace today due to technology disruptions that seem to happen on a monthly if not weekly basis. Growing companies and organizations must constantly evolve and change in order to stay relevant with customers and the marketplace. The public cloud was built to support these disruptions and allow a business to react rapidly. But the IT group and staff needs time to learn the new tools and processes, and should be empowered with the flexibility in choosing how deep these changes should go and how quickly to make the change. It is best to prepare a long runway as to not distract staff or inadvertently impact your customers.
Most articles touting the benefits of the cloud try to make the argument that a move to the cloud represents an immediate cost cutting situation from a staffing perspective, providing an immediate return on the investment. And while it’s true that a CIO can reduce OPEX for indirect staff focused on system, storage, network administration, that CIO might also offset this problem by keeping staff that have cloud specific skills in devops, and cyber security. At the end of the day, someone still needs to oversee the cloud “tenant” containers or virtual machines that are deployed. But remember those legacy custom applications that are still on-premise and critical to running a growing business? Well now the CIO has the added headache of having two technology groups to contend with, which invariably creates staff conflict, as the cloud group typically wants to force the legacy on-premise group to modernize and move up to the cloud without necessarily knowing all the business reasons around risk and impact. Instead of being able to focus on business, now the CIO has to be an employee facilitator and mediator to support change and innovation, while still providing justification for the “status quo”.
7. Disaster Recovery
I definitely like the hybrid approach when it comes to disaster recovery and business continuity. Technology will fail for whatever reason; a CIO could have high availability in place but no platform can guarantee a 100 percent uptime on it’s own. My suggestion to that CIO? If you are primarily using the cloud for an application then implement a clone or replication strategy to an on-premise in a data center co-location, if you can afford it. That gives your company the ability to failover and keep your business running in the event of a network outage. And this works vice versa if the application is primarily on-premise, replicate or clone to the cloud. Lastly, if the company is running proprietary legacy on-premise applications the CIO will want to replicate to another DR co-Lo.
According to Storage Swiss analyst George Crump, “the hybrid cloud concept gets a lot of attention and deservedly so since it is the most practical use case for many organizations. It allows them to move into cloud computing gradually and then expand as they see benefit. It also lets them return to an on-premises model or expand to multiple clouds as the needs of the business dictate.”
If you are a CIO at a growing, mid-size company or organization I would encourage you to spend an inordinate amount of time talking to your peers and taking a deeper dive into the horizontal technology media (ComputerWorld, TechTarget, CIO, InformationWeek, et al) covering the cloud, rather than the mainstream tech press titles like Techcrunch and Recode, or technology journalists at national business publications like the Wall Street Journal or The New York Times. The mainstream media’s message that almost every company should consider going all-in on the public cloud is not only misguided, but potentially harmful to those companies in growth-mode that need to scale in a cost-effective and efficient manner. Think hybrid!
This article is published as part of the IDG Contributor Network. Want to Join?