Waterfront Deployments

Deploy straight
to the field.

Ship containers to industrial PCs and servers, x86‑64 or ARM, and start them from your browser.

Fig. 1Hardware the mutexer agent runs on, by architecture. Push a multi-architecture image and each device pulls the variant for its own processor; any Linux device on these architectures will do.

Deploy without a site visit.

Any registry.

Point a program at an image on Docker Hub, Amazon ECR, Azure Container Registry or your own registry, choose the tag, and deploy.

Logins kept in Strongroom.

Registry credentials live in Strongroom Secrets Management and are resolved when a deploy runs, never pasted into a script.

From the device tree.

Every program belongs to a device. Deploy it, stop it, start it or remove it from the same tree you use to find it.

Every step, on the record.

  1. 01Sent to a live device

    A deploy goes out only when the device is online. Nothing waits silently for a machine that's gone dark.

  2. 02The right image for the processor

    Docker on the device pulls the variant that matches its architecture: x86-64, 64-bit ARM or 32-bit ARMv7.

  3. 03The agent reports in

    The device reports each step as it takes it: received, pulling, stopping the old container, starting the new one. Pull and login errors come back straight away and mark the deploy failed.

  4. 04One container per program

    Each deploy stops and replaces the previous container, so a program never runs twice on the same machine.

  5. 05Secrets at start, never in the image

    Secrets bound to the program in Strongroom become environment variables when the container is created, so they never sit in an image layer.

  6. 06Running, and staying running

    Containers restart if they exit or the device reboots. Start, stop or remove a program from the browser.

Container output, live from the device.

Open a program and read what it's saying: the latest thousand lines, then each new one a second or two after the container writes it.

Straight from Docker

The agent follows each container's output through Docker on the device. No logging library to add and no code to change.

Opens when the deploy lands

When a deployment completes, the program's output opens on its own. The first line you see is your container starting.

Picks up after a drop

If the device loses its connection, it resumes from where it left off on reconnect, as long as Docker on the device still holds the lines.

Technical specifications.

Container deployments and live output for field-deployed devices.

WorkloadContainer images
Container registriesDocker Hub, Amazon ECR, Azure Container Registry, custom
Registry credentialsStrongroom, resolved at deploy
Architecturesx86-64, ARM64, ARMv7
Deploy targetOne device per program, online at deploy
Deploy statusStep by step from the device, with failures reported
Program controlStart, stop, remove
Restart policyAlways
NetworkingHost network
SecretsStrongroom, as environment variables at container start
Container outputLatest 1,000 lines, then live
Device runtimemutexer agent
Supported devicesAny Linux device the mutexer agent runs on

Push the image.
Watch it run.

Choose an image, deploy, and read the container's first line of output moments later. In every mutexer project.