Console

The console lets you look inside a running container. It goes through the agent on the server and is not required to use Deplo.

There are two tabs:

• Shell runs one docker exec per line you type. Each line is a separate process, so cd does not carry over to the next command, and interactive programs like top, vim, less or psql do not work. Use it for ls, cat, env or a one-off script.
• Attach connects to the main process's input and output. This is a real terminal, but only works for a container started with a TTY and an open stdin.

Open a consoleCopy Anchor Link

The console is hidden until you acknowledge it once.

Pick the container from the picker. Each row shows a running dot, the service name, and an app badge on the one your domains route to. The prompt shows the user as user@service$.

The Shell select offers auto, /bin/sh, /bin/bash and raw exec with no shell. An image with no shell at all (distroless or scratch) says so. You can still run a binary that exists in it.

Good to knowCopy Anchor Link

• Changes inside a container do not survive a redeploy. The container is recreated from the image. Keep changes in a volume or in your repository.
• open_app_console and open_database_console are separate, sensitive capabilities. A console runs any code as that container's user.
• The console needs something running. A stopped app has no container to enter.
• This is not a way onto the host. You are inside a container. Reaching the server itself needs host access.

TroubleshootingCopy Anchor Link

More in Servers and agents.

Next stepsCopy Anchor Link