Containers, the comic

Every cook at their own station.

The same kitchen as the other comics, now up close. Each cook works at a station of their own: a container. Here is how you look over their shoulder, change their station, and notice when something is quietly wrong.

1 Every cook at a glance

Every cook works at a station of their own: a container. Cooks on the same dinner stand together, the containers of one compose project, and you can make groups of your own next to them.

Each cook shows how hard they are working right now: live CPU and memory per container, so the one who is sweating stands out.

Start, stop, restart or pause one cook, or a whole group at once. And for the one who will not stop: stop immediately.

2 Over a cook’s shoulder

Open a cook and everything about them has its own tab: the log, health, events, environment, processes, files, networks, and a shell.

The kitchen diary from a moment you choose: the last hour, since the start, or around the moment it went wrong. And the whole diary kept as a file.

Look in a cook’s drawers: download and upload by dragging, rename, add a folder, edit a small note. If someone changed the note in the meantime, the app says so before it saves.

3 Changing a station

Other ingredients, another door, more room: environment, ports, restart policy, volumes, health check, memory and CPU. Docker cannot change most of that on a running container, so the app builds the new station for you.

And if the new station does not work, the app puts the old one back, as it was.

A second cook just like this one, with a door of their own and a copy of their pantry: duplicate.

4 Writing it down

Turn a station, as it is right now, into a recipe others can start from: save as image. The app lists what does not come along, such as the pantry.

Or write down how this cook was hired: the docker run command, or the same as a compose service to put on the recipe card.

Something odd to report? One click copies a diagnosis, ready to paste into an issue, with the secrets left out.

5 The app notices

A cook can keep cooking after the hatch was bricked up and a new one opened further along: the network was removed and created again. The container still runs, but nothing gets through. The app notices, and offers the fix.

Or a cook who started while their door was already taken: running, but nobody comes in. The app says who is standing at that door.

In Docker Client MX: the Containers page, grouped the way you work, and a page of its own for every cook.

The cheat sheet

Every picture above, in Docker words.

A cook at a station: A container
Cooks on the same dinner: The containers of one compose project
How hard a cook works: Live CPU and memory
A break for the whole group: A bulk action
The cook who will not stop: Stop immediately (kill)
The kitchen diary: The logs
A cook’s drawers: The files in a container
A new station, same cook: Recreating with new settings
A second cook, own door, own pantry: Duplicate
The station as a recipe: Save as image (commit)
How the cook was hired: The docker run command
The hatch that moved: A network that was created again
Someone else at your door: A port that was already taken

Now look over your own cooks’ shoulders.

Every container grouped the way you work, with its logs, files, health and a shell one click away, and changes that put the old container back when they fail.

What the app does with containers →Compose, the comic →Download for Windows