Codafords runs app hosting, functions, databases and storage on a private, self-managed cluster we own end to end.
Five purpose-built layers. Traffic flows top to bottom; a tenant’s code flows bottom to top.
Every route enters through the ingress. TLS certificates are issued automatically, for platform URLs and custom domains alike.
The platform API, scheduler and identity run in their own pool, dedicated to Codafords.
Workloads are split into isolated pools. A cluster policy enforces node affinity, so tenant workloads run only in the tenant pool, physically separate from platform infrastructure.
Block storage for anything that needs a low-latency disk, object storage for everything better kept as blobs.
Each project is its own quota’d, isolated workload. NetworkPolicies restrict traffic across tenants, and HTTP workloads scale to zero when idle.
Event-driven autoscaling drives tenant apps and CodaFunctions, and can scale HTTP workloads to zero.
An admission-time check verifies cluster headroom before a new tenant is provisioned. Pilot stage.
Every credential generated in the project is logged in one access-controlled, git-ignored record.
A GitHub repo through a GitHub App, or a push to Codafords GitLab.
Build detection with no Dockerfile, plus a custom Next.js buildpack.
Built on the DevOps pool, isolated from tenant workloads.
Image stored in the platform’s own registry, scoped per tenant.
Rolled out into the tenant pool with TLS and live status.
Every product below has been run end to end on the running cluster.
Push code, get a running, publicly reachable app.
Deploy a single function instead of a whole app.
A dedicated production-grade database per tenant, provisioned in minutes.
An S3-like storage service, API-driven by design.
One dashboard for everything a tenant runs.
Every layer above is running right now. Connect a repository and get a live URL in about ninety seconds.