What are you looking for ?

Giant Swarm Leverages Quobyte Storage Software

To run and automate storage for micro-services infrastructure

Running storage inside containers: Giant Swarm GmbH leverages Quobyte, Inc. software storage capabilities to run and automate a reliable storage basis for their micro-services infrastructure.

Quobyte,Giant Swarm

Container experts
Giant Swarm is a micro-service infrastructure provider based in Cologne, Germany. Feeling the growing pains when trying to scale a monolithic app during a former project, the three founders set out to find a way to handle the scalability problem: isolate problems into small services. Their resulting un-opinionated solution for micro-service infrastructures blends existing open-source components – choosing CoreOS as a basis – with custom-created software and third-party tools. Together, this makes for a framework for running micro-services.

The company is expanding its portfolio after open-sourcing their technology stack to help businesses tailor container-based infrastructures to their needs. Serving enterprise clients on premise, equipping SMBs with a managed solution, and procuring the infrastructure in their own data centers, they provide fast and reliable services and help their clients set up a scalable and flexible container infrastructure.

Flexible storage for containers
In order to find the right storage system that would fulfill all of Giant Swarm’s needs and provide a foundation for their micro-services framework, the team decided to first evaluate various software storage systems. The guiding idea was to find a solution that just worked and would let them focus on developing their Container as a Service (CaaS) platform instead of adding overhead to managing the infrastructure. Early on, they focused on Ceph and Quobyte because the two had the most promising features that would fit their needs: a shared file storage interface and massive scalability.

Containerizing storage becomes matter of plug-and-play
With both products checking those two features off the list, one distinguishing factor in Giant Swarm’s decision making was to find a reliable and scalable storage solution that could run inside containers. Containers have so far not always played well with storage: they are great for virtualizing stateless applications but when a containerized application requires file system access, it becomes a lot harder to move the container from one host to the next because the container’s data volume is tied to a specific server. As a distributed file system, Quobyte has the solution baked right into its core: it makes mounting a storage volume inside a container a matter of plug-and-play.

Migrating containers within live system
A related benefit was the possibility of live migrations: whenever you move a container to another host, its attached volume is already there. If a host fails and the container shuts down, you can fire it up on another host and reattach the storage volume.

Running complete storage infrastructure inside container
What made Quobyte valuable for Giant Swarm was the fact that Quobyte can be run inside a container. Giant Swarm thus benefits from a one-click agility: they could deploy the complete Quobyte storage infrastructure with the ease of deploying a container.

That’s when Quobyte really tipped the scales,” said Dennis Benkert, lead storage administrator, Giant Swarm, “their product was the only one with the capability to run a complete storage infrastructure inside a container. That’s a whole new level when it comes to ease of deployment!

Fitting in with container ecosystem
No less important were the question of overall compatibility with the container ecosystem in general and the performance issue: is storage access fast and reliable enough for Giant Swarm’s needs? Quobyte came out strong – in both respects. As far as integration with current tools go, Quobyte integrates with orchestration tools like Apache Mesos, Mesosphere DCOS, and Kubernetes. And as for performance: the testers were thrilled with the IO/s provided.

Support is more than lip service
Giant Swarm decided to subscribe to Quobyte because of its simplicity, functionality, and performance. But an important plus weighing in for Quobyte was that the developers at Giant Swarm could rely on Quobyte’s first-rate support.

I could call in and talk with a developer directly and have my problem taken care of asap. That also really set Quobyte apart and made the whole experience even more worthwhile,” Benkert said.

Playing well with Giant Swarm’s hyper-converged setup
Giant Swarm deployed Quobyte on their hyperconverged server infrastructure and integrated the container lifecycle with Quobyte’s volume management. When customers start up a container, a corresponding Quobyte volume is created and attached to the container. Moving a container between hosts leaves the container-volume relation intact.

Future expansion and scaling services without downtime
With Quobyte, Giant Swarm deployed a drop-in file system that could be used with a range of container workloads. Quobyte’s extensive API made the integration with their container lifecycle an easy task. In day-to-day operations, their Quobyte storage stays out of the way and keeps storage up across server reboots and hardware failures – without any downtime.

Quobyte positions Giant Swarm for future expansion.

Looking ahead, we’re all set for whatever our clients decide. Whether they want to switch from HDDs to SSDs or need to quickly scale out due to temporarily high traffic, we don’t have to go out of our way to make it happen,” Benkert says. “Also, thanks to Quobyte, we can now easily replace old machines without any downtime.

Reflecting on the ongoing experience, Timo Derstappen, Giant Swarm’s co-founder and CTO, adds: “I really like Quobyte’s solution. It exceeds our already high expectations in terms of storage performance and we’re thrilled about how smooth it runs. In our case, it’s of course also a major benefit that we could seamlessly incorporate it into our micro-services framework.


  • Run a complete storage infrastructure inside a container
  • No downtime with rolling updates
  • Scale storage up and down flexibly and depending on business needs
  • Performance for business-critical databases
  • Increase reliability for data safety and lights-out operations
  • Run storage DevOps-style
  • Reduce maintenance time and administrative expenses