How We Speed Up Deployments By 30%
The Essence of Container Image Mirroring
When a cluster is deployed on a cloud account, Qovery creates a dedicated image registry, which essentially acts as a mirroring system. This system ensures that every time an application is set to be deployed on your cluster, the application image is duplicated onto the mirroring registry.
Why is this essential? Image mirroring is a paramount best practice. Envision a scenario where an application runs on a production environment, and Kubernetes needs to pull the image again to launch a new instance. It would be disastrous if this process failed due to the unavailability of the source container registry. By ensuring there's always a copy on the container registry adjacent to the Kubernetes cluster, we eliminate this risk.
How Does Image Mirroring Work?
For Applications Deployed from a Git repository
Before initiating the build for an application, say application A1, Qovery first checks within the mirroring registry for a previously built version. If the image, complete with the same version (determined by commit ID and environment variables), is found, the build process is skipped, and deployment on the Kubernetes cluster begins.
For Applications Deployed from a Container Registry
The behavior in this case hinges on the selected mirroring mode within the cluster's advanced settings.
Two modes are available:
Service Mode: In this default mode, images in the mirroring registry are grouped by "Qovery service", ensuring each service's mirroring process is entirely separate. If an image, marked by its unique name and tag, exists, the mirroring is overlooked, and deployment begins. However, if the same image is employed across varying environments or services, it will be mirrored repeatedly, which may decrease deployment velocity.
Cluster Mode: Notably unavailable on Scaleway, this mode means all applications on the same cluster are mirrored within the same repository. If an image, bearing the same name and tag, is found, mirroring is skipped, and deployment commences. The highlight? If an identical image is used across environments or services, the setup refrains from duplicating the mirror process, hence hastening deployment.
Unlocking 30% Faster Deployments
With the streamlined processes outlined above and our commitment to ensuring the utmost efficiency, we've achieved a 30% enhancement in deployment times. Every decision, from using unique image tags to selecting the right mirroring mode, plays a crucial role in this improvement.
It's worth noting that when working with containerized applications, having unique image tags is non-negotiable for accurate version management. Failing to employ unique tags can lead to complications due to caching mechanisms, not just by our mirroring system but also by Kubernetes. Ensuring this has been instrumental in achieving our accelerated deployment speeds.
Read the complete documentation of this feature here.
Any thoughts? Feel free to reach out on our forum.
.svg)
.svg)
.svg)
Suggested articles
Scaling your deployments to zero is only half the battle. If your cluster autoscaler does not aggressively bin-pack and terminate the underlying worker nodes, you are still paying for idle metal. True environment sleeping requires tight integration between your ingress layer and your node provisioner to actually realize FinOps savings.
The biggest mistake enterprises make when evaluating Kubernetes management platforms is confusing infrastructure provisioning with Day-2 operations. Tools like Terraform or kOps are excellent for spinning up the underlying EC2 instances and networking, but they do absolutely nothing to prevent configuration drift, automate certificate rotation, or right-size your idle workloads once the cluster is actually running.
For years, Red Hat OpenShift has been the safe choice for heavily regulated, on-premise environments. It operates as a secure fortress. But in the public cloud, that fortress acts as an expensive prison. Paying proprietary per-core licensing fees on top of your standard AWS or GCP compute bill is a redundant "middleware tax." Escaping OpenShift requires decoupling your infrastructure from your developer experience by running standard, vanilla Kubernetes paired with an agentic control plane.
Agentic Kubernetes resource reclamation is the practice of using an autonomous control plane to continuously identify, suspend, and delete idle infrastructure across a multi-cloud Kubernetes fleet. It replaces manual cleanup and reactive autoscaling with intent-based policies that act on business state, eliminating the configuration drift and cloud waste typical of unmanaged fleets.
Kubernetes focuses on container orchestration, but the reality on the ground is far less forgiving. Provisioning a single cluster is a trivial Day-1 exercise. The true operational nightmare begins on Day 2. Teams that treat multi-cloud fleets like isolated pets inevitably face crushing YAML configuration drift, runaway AWS bills, and severe scaling bottlenecks.
Rancher solved the Day-1 problem of launching clusters across disparate bare-metal environments. But in 2026, launching clusters is no longer the bottleneck. The real failure point is Day-2: managing the operational chaos, security patching, and configuration drift on top of them. Rancher is a heavy, ops-focused fleet manager that completely ignores the application developer. If your goal is developer velocity and automated FinOps, you must graduate from basic fleet management to an intent-based Kubernetes Management Platform like Qovery.
The shift from AI copilots to autonomous agents is redefining infrastructure requirements. Discover how to build secure, stateful, and compliant Agentic AI systems using Kubernetes, sandboxing, and observability while meeting EU AI Act standards
.webp)