Compute abstractions
Run anything, anywhere, uniformly
Containr workloads run the same way whether they're on a Kubernetes cluster in the cloud, a vessel at sea, or an edge site at the port. Services, jobs, batch pipelines, and streams all speak the same language.
Workload types
Every compute pattern in one model
Define services, jobs, pipelines, and streams using the same manifest syntax. Containr handles the orchestration.
Services
Long-running HTTP and gRPC services with automatic load balancing, health checks, and instant failover.
Jobs
One-time or scheduled tasks that run to completion, with built-in retry logic and result tracking.
Batch pipelines
Multi-step data processing pipelines that scale parallelism automatically based on workload size.
Streaming workloads
Continuous event processors and message consumers with exactly-once semantics and dead-letter handling.
Why workloads matter
Uniform logs, metrics, and rollback across every workload type
No more translating between Kubernetes Job syntax, ECS task definitions, and systemd units. Same manifest, same CLI, same guarantees.
Single manifest
Define all workload behavior in one YAML file, no dialect switching.
Instant rollback
One command reverts any workload to its previous version, anywhere.
Queryable logs
Every workload's output flows into structured logs you can search and aggregate.
Ready to ship on Containr?
Talk to our crew about standardizing your container infrastructure — from the yard to the cloud.