Public Cloud now offering the Abracadabra Package!
Seems like every other week we hear about a new service coming out of the public cloud: Lambda! Compute Engine! BigQuery! Glue! Athena! Fargate! Purview!
Each of these new services comes with a new innovative marketing campaign, unique name, and the feeling that this is going to be better than ever. This reminds me of when Apple introduced 5G on their phones, this already existed on Android for a while, as a new innovative feature!
An important item to note, is that most of these public cloud services exist in some capacity as a stand-alone service. If you pair the stand-alone service to a clustered highly available environment, you have a service match. So what exactly do you need?
Many articles online compare these “new innovative” services offered by big tech giants against each other, implying that there are no alternatives. But that’s not true, at least not anymore.
OpenStack has thousands of developers working on various cloud-oriented projects to provide the same services these public cloud providers offer. We’re only going to cover a couple of them here, but it should give you an idea of what to look for when looking at alternatives.
In this article:
Database as a Service (DBaaS)
Database as a Service aims to provide a highly available, large volume, performant database for your applications to plug into. This isn’t a new concept and comes in a relational or non relational format. Azure calls their product SQL Database, Google calls their product Cloud SQL, AWS calls it Aurora or Couchbase.
OpenStack has a project called Trove, the main focus of which is to provide scalable and reliable DBaaS for various database engines and formats. It’s an OpenStack technical project that you can add to your OpenStack cloud at no additional cost.
Now your private cloud has the same service the big cloud providers have in their marketing playbook to lure database operators and make extra dollars from the same infrastructure.
Functions as a Service (FaaS)
You may not be familiar with Functions as a Service yet, but it’s catching on quite fast. The idea of Functions as a Service is to send a compute workload elsewhere and get a returned result.
Many image processing applications will shoot off the user uploaded image for resizing, formatting, or AI manipulation to a function service running elsewhere. This gets the image back to the desired format, size, or compression leveraged result. It’s becoming common as we approach the split of micro-services to delegate specific tasks to a specific component not local to the environment running the application.
This simple service is relabeled and sold as an add-on product by the house. Azure Functions, AWS Lambda, and Google’s Cloud Functions all tackle the principle of offloading a specific task yielding a desired compute-driven result.
In Openstack, this functionality is known to exist under the Qinling project. The goal of the project is to provide these “serverless” functions using various technologies like Docker and Kubernetes backed by the support of different storage mechanisms. Once again, you are able to run this on your OpenStack private cloud as an added value service, without it being an extra item added to your public cloud bill.
What about something as simple as cloud object storage? This technology came out awhile ago. It allows you to combine the pieces of a file with some metadata, a custom identifier, and store it on a highly reliable, available, and performant storage locale. File system sharing and mounting across a bunch of servers to increase file availability doesn’t exist anymore. Just ask the service to give you your file by the identifier and you’re good to go! Start organizing your data and files into buckets – basically sectors – and you can have a highly leveraged storage behind your data delivery necessities.
You could run an application like MinIO in a docker container and get the same object storage functionality as marketed by AWS S3, Google Cloud Storage, or Azure Storage. However, that would just be a stand-alone service with many, many limitations.
The selling point of the public cloud offering is the high availability and data redundancy, right? You can set this up on your own OpenStack cloud! It integrates this functionality under the Swift service project, so you can store and retrieve lots of data using a simple API. You’re able to plug Swift into a variety of backends like Ceph – another resilient storage technology – and be on your way to utilizing the latest and greatest storage mechanics.
Do your Due Diligence
There are dozens of OpenStack projects that compete with the everyday public cloud offerings of tech giants that can run on your own private cloud.
The next time you hear someone preach about some “marketing new name service” offered by these market majority holders, I strongly urge you to research what those services truly mean. If they are important, are they available in some other way than the marketing jargon you are being fed?
I wouldn’t be doing this write up justice if I didn’t mention that the entire point of Flex Metal OpenStack Private Cloud by InMotion Hosting is to make OpenStack private clouds accessible to anyone with a much lower barrier to entry than ever known before.
Once you have your private cloud – you can set up anything you need on it that these big guys offer. You can get the benefit of these latest cloud technologies without having to depend on the extra nickels and dimes they’re going to charge you for these ‘Extras’. Just do your due diligence.
If you want to learn more about Private Cloud and OpenStack check out some of our articles: