Boosting performance and cutting costs:
Optios' journey to Google Cloud
In de oktobereditie van het TDSynnex magazine werd ruim plaats gemaakt voor een interview met Sven Cornelis (Optios) en Roel Van Steenberghe (vBridge). In het artikel worden de ingrediënten opgelijst van de succesvolle samenwerking voor de migratie van hun SAAS-platform naar Google Cloud.
De volledige tekst vind je onder de afbeelding (EN).
Optios is a Belgian scale-up that offers hairdressers and beauty salons software to let customers do bookings online, and that also manages the cash register, stock keeping, accounting…
Initially developed as a student project, the company now has over 3.000 customers in Belgium, Luxemburg and the Netherlands and pursues a growth strategy to conquer other neighboring countries.
Recently, Optios embarked on a migration to Google Cloud, assisted by the cloud architecture specialists from vBridge. To know more about the project, we talked to Sven Cornelis, Co-founder of Optios and Roel Van Steenberghe, founder of vBridge.
A reliable and robust platform
Sven, as a SaaS business, Optios had already been using a cloud platform for some years. What made you want to switch to another cloud platform?
Sven Cornelis: We experienced that the costs of the other platform increased exponentially as our database grew. Every time we wanted to add extra CPUs, the cost doubled. That’s a huge disadvantage if you want to grow your business. The platform lacked flexibility and robustness. It felt like we had reached the limits of what we could achieve with it. As an example, if we wanted to roll back a feature we had deployed because it contained a bug, that could take up to an hour. That was unacceptable to us.
What were the requirements you had that led you to choose Google Cloud?
Sven: After the experiences we had, we wanted a platform that was reliable and robust and that offered performance at a more reasonable price point. Also, we knew that we wanted to start using Kubernetes. As Kubernetes was developed based on the internal container orchestrator of Google, and they are still continuing to evolve Kubernetes, we knew they have an advantage over other platforms. It’s part of their Anthos platform, of which we wanted to use different features. Having the possibility of making a connection to Google’s analytical features was another good reason to go for Google Cloud. In summary, the decision for Google Cloud was based on the feature set and the combination with other services. And the lower cost for scaling our business. Our solution architect had previous experience with vBridge for cloud implementations. It was important for us that vBridge is about the same size as Optios. That creates an understanding. Fun fact: Roel Van Steenberghe of vBridge was also our professor at the University of Applied Sciences where we studied IT.
Technical implementation
It’s a small world, as the saying goes. Roel, can you tell us a bit more about the technical aspects of the implementation. Sven already mentioned Kubernetes, but I guess there’s more?
Roel Van Steenberghe: Google Kubernetes Engine is the basis of the entire infrastructure, that’s where all applications run. And Kubernetes also allows for a big variety of services to be added. Another important factor in the set-up at Optios is the Cloud SQL, where the database runs. And let’s not forget Service Mesh. Kubernetes uses containers, and as the environment grows, you get lots of containers and it’s easy to lose the overview over all these containers and the interactions between the containers. Service Mesh keeps that overview. Without Service Mesh, it’s difficult to find potential bottlenecks when there is a performance issue. What’s more, Service Mesh makes it easier to define what containers need to communicate or should not communicate with each other. This is a good way to improve security.
Migration as a combined effort
How did the implementation go? Who was responsible for what part?
Sven: It was very much a combined effort. Our own solution architect was responsible for the design of the architecture, and we managed the overall planning and the communication between vBridge and our internal services, to ensure the necessary changes were made to the applications to prepare them for the migration. We also worked together on defining the environment to run our applications. We see vBridge as a partner that we can use as a sounding board for our ideas. They always find a solution when we are confronted with a specific challenge. They have a good relationship with both Google, which gives them access to the right specialists, and they can also rely on support from the cloud engineers at TD Synnex. vBridge documented everything they did and transferred a lot of knowledge to our own team, through transfer meeting and regular stand-up meetings. This empowers us to do a lot more ourselves.
How seamless was the migration to the customers of Optios?
Roel: We took a gradual approach to the migration, not a big bang. The customers probably only noticed performance increases, so we did not get any complaints about that. Everything really went smoothly. A migration always comes with risks, of course. So you need to work in a structured and systematic way. Optios already used containers, which made it easier to migrate each container separately. This enabled us to work quite quickly and not have two cloud services running in parallel for too long.
Sven: We started by migrating our own internal services. That allowed us to develop procedures and fine-tune them on our own services. Next up were the less complex external services, which gave us the necessary experience and comfort to tackle the most complex services. Downtime was very limited and we planned this during the night.
In all, what would you say are the most important results after the migration?
Sven: The performance of our applications is a lot better than before, and the load times are much shorter. We also found that management of the cloud environment became easier, saving a lot of time. The entire DevOps cycle is more structured now, among other things by the use of Infrastructure-as-Code. The migration also enhances the security of our environment. Security was something that we designed into the project early on. And we are also sure that the cost will no longer increase exponentially as we grow our business. Thanks to Google Cloud, we are now future-proof. The implementation will serve us for many years to come and provides the flexibility to cope with anything the future brings.