Public clouds have become popular as a solution for many data-driven projects around Big Data, Business Intelligence and Analytics, as well as specialized projects, such as Machine Learning and AI. Cloud providers normally offer significant scalability and advantageous, user-friendly tools. Migration to the cloud is considered to be cost-effective, making it an excellent choice for businesses and organizations of all sizes. However, the use of public clouds come with a number of challenges, and these should be considered before embarking on such a venture.
Learning to work within a cloud environment, and the even more complex multi-cloud landscape now becoming more widespread, would preferably be a quick and easy experience. That is often not the case though. Not too surprisingly, when working with some of the biggest cloud vendors like AWS, Azure, and Google, it isn’t quite as simple as they suggest it should be. Public clouds offer tens, if not hundreds of different services, and a person needs training to use those services (larger businesses often use individuals with skills specific to the cloud vendor). Ideally, training takes place in a “free” setting, and several clouds have begun offering a month to a year of free services to learn their system.
- AWS offers unlimited access to over 100 courses.
- Microsoft Azure offers 12 months of free services.
- The Google Cloud Platform offers 12 months free, with $300 in credit for GCP services.
Efficiency and Cost
ParkMyCloud helps enterprises to automatically identify and eliminate unnecessary cloud expenses. The cloud is often charged like a utility, said Jay Chapel, the founder and CEO, in a recent DATAVERSITY® interview, and ParkMyCloud helps to turn the lights off when they’re not needed. It also helps people automatically resize their resources. It is quite common for organizations to procure a resource because they think it is the size they need, and often it is much larger than is actually required. “Companies need data and automated action on cost effectiveness and cloud resource management,” said Chapel.
“We ‘plug’ into each of the cloud provider’s APIs, and we pull out metric data from those APIs. That’s usually things like CPU, disk, network, and memory. We pull that data out into our platform, and then we baseline that data, and then make recommendations to people to either schedule things based on idle time, or re-size things based on lack of usage.”
Security concerns commonly associated with the cloud fall into two general categories remarked Chapel: security issues faced by cloud providers and security issues faced by their customers. While the cloud provider is responsible for infrastructure security, including their clients’ data and applications, it must also be noted “the user” has some responsibilities, as well, and should take steps to use strong passwords and identification measures. In regard to cloud security issues, Chapel commented:
“Security is either the highest priority, or the second highest priority, when you talk to people about cloud computing. Security is one and cost is two, or vice versa, and I don’t think that’s going to go away. So, we’ve built a platform that overcomes those security concerns. We don’t deploy agents, we pull everything through an API, making us pretty low risk in that regard. And the way we do a handshake with them, through their accounts, is again, pretty low risk.”
An important aspect of planning for a cloud deployment involves predicting your monthly costs. In calculating monthly cloud expenses, all aspects of the data going to the cloud, and the projects being planned, should be considered. By breaking down the costs by each feature that will be needed, a more accurate budget can be developed. Having open, in-depth discussions with staff and cloud reps about business and technical requirements can be very beneficial.
On the topic of saving money, Chapel said that “ParkMyCloud really helps people with automated cost control. We help people schedule resources to turn off during nights and weekends.” The cloud is pretty much set up with a “the light is on all the time until you turn it off” philosophy, he commented. “We help people turn the lights off when they’re not there, and then we help people automatically resize their resources.”
Cloud Data Governance
Cloud Data Governance uses a set of rules that are created, monitored, and amended as needed, as a means of controlling costs, improving efficiency, and eliminating security risks. According to Chapel:
“The other area we talk to people a lot about is really governance and control, like who has access to what within a company and who can deploy new resources in the cloud, who can turn off new resources in the cloud and that type of stuff, and then of course there’s the ongoing security dilemmas. Nobody wants a breach.”
Within the cloud, different departments often develop their own systems. This can affect operational costs and can quickly get out of hand with no controls in place. Additionally, software, applications, and programs used by one department may have problems interacting with applications used by another department. A lack of control/governance creates chaos and cost inefficiencies, as well as security concerns. Compliance rules must be developed to maintain a smooth integration of all digital devices.
Compliance rules need to be monitored. Humans often come up with interesting, unplanned behavior. The goal of compliance can be achieved with the help of cloud management software, such as ParkMyCloud.
Hybrid and multi-cloud strategies are gaining in popularity. While these strategies offer benefits, using more than one cloud vendor makes Data Governance, cloud management, and cost efficiency more difficult. Businesses adopt hybrid and multi-cloud strategies to handle a variety of cloud computing projects. By eclectically taking advantage of the different tools and features offered by different clouds, an organization can maximize their benefits. Should cloud-oriented businesses specialize in specific clouds? Chapel remarked that:
“Over the last 18 months, we’ve seen a growing influx of customers that use multi-cloud. They either use AWS and Azure or AWS and Google, I don’t think we see a lot of Azure and Google users. AWS and Azure’s the most common because a lot of these large AWS customers also have Enterprise EAs with Microsoft, and Microsoft \ has been pumping Azure into those enterprise agreements.”
In most large enterprises they hire different people that manage AWS and different people that manage Azure, said Chapel, and different people that manage Google. They are different experts for different providers, and while they may be on the same team and report to the same manager, it takes specific skill sets to work with each system.
ParkMyCloud offers a way to effectively manage cloud resources, minimize problems, and save money. As an organization shifts to using a multi-cloud environment, effective cloud management inevitably becomes a priority. Cost controls can be time consuming and, ideally, should be automated and constantly monitored. If Data Governance can be included with the automated services, even better. Chapel ended by saying:
“We’re really focused on the identification of idle and oversized resources, the governance around that, and automatically helping companies manage those resources to reduce costs.”
Image used under license from Shutterstock.com