Hello again. It’s March, and while I missed February for putting out an “Explained” blog, I have a topic today that I think could use a non-developer breakdown. Let’s just get it out there, cloud computing is a term that has been so saturated and used in so many different, confusing, and sometimes incorrect contexts, that it could be considered a buzzword and nothing more. Add hybrid in front of that and again, some of you might be doing a big eye roll at this post because hybrid cloud has ALSO become a buzzword. What do you think hybrid cloud means? Can you summarize it in one or two sentences? Hybrid cloud is a blend of public and private (on-premise) cloud. Boom- article done? Not quite. Read on for a deeper look into the equivocal term, hybrid cloud.
Stratus, Cumulus, Nimbostratus oh my! No, no, I know we aren’t talking about those cloud types. Although I will note my favorite type of cloud would have to be the Kelvin-Helmholtz.https://scied.ucar.edu/image/kelvin-helmholtz-clouds
So just like the clouds in the sky, there are different types of cloud computing. I’ll start us out with some easy-to-read bullets:
- Public Cloud- So much space for activities! Public cloud is great for general purpose use cases or ones that require a lot of space for storage without a crazy high cost.
- Private Cloud- Usually operated by the company that is using it, private cloud can also be known as on-premise (on-prem, for short) and is fully managed by the provider/company for singular use. On-prem arguably has a higher capability to be more secure (if you have the right DevOps/Security folks) because, well, it’s private and no one else has access.
- Hybrid Cloud- The combination of private and public cloud, kinda like a Limequat, a combination of lime and kumquat. It's two great products put together to make something that satisfies two types of taste buds, or in hybrid cloud’s case, cloud buds 😆 We’ll go deeper on hybrid cloud later because this blog is after all about…hybrid cloud.
- Community Cloud- Think cloud, but make it a neighborhood watch. A group of organizations with similar product offerings or in similar industries come together and share an infrastructure that can be either internally managed or managed by a 3rd party. It’s a cloud block party!
Then there are different types of services that you can find in Cloud Computing:
- Infrastructure as a Service (IaaS)- An example would be AWS or Microsoft Azure
- Software as a Service (SaaS)- An example we all know (and love / sometimes hate) is Slack (does anyone else have nightmares about the Slack ping?)
- Platform as a Service (PaaS)- Examples include Heroku or Elastic Beanstalk
- Function as a Service (FaaS)- All the big-time folks like AWS, Microsoft, etc. have function as a service. A function as a service in short is serverless as a service IMO. You can define a task(function) without building infrastructure which can be a huge, costly, time lengthy job.
- Database as a Service (DBaaS)- Umm hello! HarperDB has a DBaaS offering!
Whoa, that’s a lot of acronyms! Cloud companies provide these packages or service models so that you can select what works for your use case or business model with minimal work on your end to get up and running. “As a Service” means your cloud provider is like your cloud waiter. They work with their team to provide you a clean table, delish meal, and clean up after you leave. Only for cloud computing that means they handle the infrastructure, the support, and make cloud computing so tidy that you might forget about all that is happening in the background.
When I was doing my research for this blog, and this particular section, the first article I found was titled “The Ironic History of Hybrid Cloud.” If that doesn’t spark your interest in a history lesson, I don’t know what will! While cloud computing seems like a fairly new concept (by new I mean the 2000s), it actually dates back to the 1960s which blows my mind! A book called “Intergalactic Computer Network” by Joseph Carl Robnett Licklider helped create the idea that is cloud computing as we know it today. Enter Salesforce in 1999, where cloud computing becomes an actual thing that people could use, and then we see AWS emerge in 2003 and the rest is history.
As with all good things, they usually have a downside and cloud computing was no different. Developers in larger companies were finding it harder and harder to achieve their goals with costly, time-consuming private cloud implementations. They started to use a back door option that meant skipping around their IT department and using cloud providers like AWS or managed software like Salesforce, to implement their use cases at half the cost and in half the time. The problem with this was it was breaking protocols around security, and to be fair, simply wasn’t as secure as on-prem practices. This is how we come to meet hybrid cloud. The best of both worlds: security and ownership from on-prem with the flexibility, cost savings, and endless opportunities of public cloud.
First, the nitty-gritty of what a hybrid cloud is. As you know, I love pulling from past blogs from my fellow HarperDB teammates, and this is no exception. Kyle, our CTO, calls hybrid cloud the “peanut butter to your chocolate” and you know I love a good food reference. Hybrid cloud is a combination of at least one private and one public cloud. The multiple clouds are managed by a managed service, container, API, or even a data lake! You can either connect the clouds themselves or focus on making apps that run inside each cloud portable so that they can be shifted from one to the other quickly and easily. You can really build out any blend of technologies to free up your dev and IT teams using hybrid cloud, the possibilities are endless. One example would be a development team that runs applications across both private and public clouds for redundancy and they put different software in different places for security.
Hybrid cloud is great for disaster recovery. This means that an IT team or dev team can create an exact replica of their on-prem cloud in the public cloud for failover and backup. Hybrid cloud can also be used for application development and testing environments. Public cloud is great because it's quick to get up and running, super-fast, and also easy to use, but when it comes time to deploy a lot of companies want the security of on-prem for production data. Hybrid cloud allows you to shift between the two. A third application of hybrid cloud goes by the term “cloud bursting.” This is when companies use public cloud for overflow of data in times of high activity. Think of a social media company that might monitor Tweets for a client, but when the client has a big event and they have more than normal amount of tweets coming in, cloud bursting could be used for the higher volume of Tweets they need to monitor during that event. Finally, a fourth use case might be when a company has mixed data sets that they are working with and analyzing. For example, say that same social media company has a healthcare client whose social data, for some reason, falls under HIPAA requirements or maybe they have a Europe-based client that needs to follow GDPR. They probably want to store that data in a secure private cloud while the event client’s tweets can be stored in the public cloud.
Enter multi-cloud. Multi-cloud is two or more public clouds or private clouds, or a combination of many public and private clouds. Basically, it’s a lot of clouds ☁️ ☁️ ☁️ ☁️ . Multi-cloud means that these public, private, or hybrid clouds are connected in some way. Multi-cloud usually means an organization has very specific use cases for their different types of clouds. They also may use this strategy to minimize downtime, data loss, and create more storage for their organization.
What type of cloud model do you use for your projects or at your company? What is the best option in your opinion, and do you have a cloud provider you are a fanboy/fangirl/fanperson of? I love to hear from you so let me know! HarperDB (of course!) has my vote as the best hybrid cloud solution, but we all know I am biased 😉 We are constantly rolling out new features and functionality that make the product ideal for low latency, high-performance data ingest, and data analytics.
I’m always taking topic suggestions, you can DM me on Twitter with your ideas or feedback if commenting isn’t your thing: @kaylanne04!