Skip to content
laraportlaraportlaraport

Container list

Opening a server shows all its containers, running and stopped.

The server panel

Section titled “The server panel”

Above the list, a panel shows the state of the server itself:

╭──────────────────────────────────────────────────────────────────────────────────╮
│ ● Production    Ubuntu 24.04.1 LTS · 6.8.0-45-generic · x86_64    up 5w · prod-1 │
│ Load     ━───────── 0.42 0.51 0.48  ·  8 cores                                   │
│ Memory   ━━━━━━──── 10G / 16G  ·  no swap                                        │
│ Disk     / ━━━━━━━━━━ 76G / 80G, 4.0G free · docker                              │
│ Docker   27.3.1  ·  overlay2  ·  Swarm manager, 1 node  ·  Dokploy               │
│ ⚠ disk / is 95% full, 4.0G free  ·  ⚠ reboot required                            │
╰──────────────────────────────────────────────────────────────────────────────────╯
  • Load is shown against the number of cores: a load of 8 on 8 cores is a full machine.
  • Memory counts what is really available to programs (MemAvailable), not just free memory.
  • Disk shows the root filesystem and, when it lives elsewhere, Docker’s data directory. It turns yellow from 85% and red from 95%. To free space, press i (images, volumes and networks) or g (cleanup).
  • Warnings appear for a nearly full disk, low memory, a load above the number of cores, a pending reboot (/var/run/reboot-required) and a Swarm in trouble.

The panel is read with one command that changes nothing on the server: /etc/os-release, /proc, df and docker info. It is read again with r.

p switches between the full panel and a single line, and remembers the choice. On a short or narrow window the panel shrinks to one line by itself, so the list keeps room. Settings (s on the server list) can also switch it off: then the server is not queried at all.

When Docker runs in a virtual machine (Docker Desktop or OrbStack on a Mac) or on another machine (docker -H, a Docker context), the panel shows what Docker reports, cores and memory, but not load and disks: the numbers of the machine laraport is connected to would describe a different system.

What each row shows

Section titled “What each row shows”
  • The name, or the display name you gave it.
  • A compact status: 3mo ✓ for a container up three months and healthy, exit 137 · 8h for one that was killed 8 hours ago. Health is shown by colour.
  • Published ports.
  • Live CPU and memory, when the window is wide enough (see usage_columns in Configuration).

Groups

Section titled “Groups”

Containers are grouped by project:

  • a Swarm stack;
  • a Coolify project, from the labels Coolify puts on its containers, with the environment when it is not production (shop/staging), and Coolify’s own containers in a coolify group;
  • a Dokploy project, from Dokploy’s <project>-<app>-<random> service names, with Dokploy’s own services in a dokploy group;
  • a Compose project;
  • or the group you set for a container.

Coolify containers are listed by their resource’s name (api, shop-db) instead of their UUID name, with the service for a resource of several containers (stack/worker) and the pull request for a preview (api (PR 7)).

Containers in a group are indented under its header. enter or space on a header collapses the group; v switches grouping off. l on a header follows the logs of every running container of the group in one stream (see Logs).

Old tasks and hidden containers

Section titled “Old tasks and hidden containers”

Every Swarm or Dokploy redeploy leaves stopped tasks behind. When a newer task of the same service exists, the old ones collapse into one line (· 4 old tasks). . shows them, and the containers you hid.

Keys

Section titled “Keys”
Key Action
enter Open the container
/ Filter
l Logs; on a group header, the logs of the group
t Shell in the container
f File browser
s / x Start / stop
R Restart
D Remove
e Container settings: display name, favourite, hidden, root
. Show hidden containers and old tasks
v Grouping on or off
i Images, volumes and networks of the server
E Events of the server’s containers in the last 24 hours
g Server-wide cleanup commands
p Server panel: full or one line
r Read the list again

Stop, restart and remove ask first, with Cancel focused. The list reloads afterwards and shows the result.

For a Swarm task, the confirmation explains that Swarm starts a replacement. To restart the whole service, open the container and use Redeploy service on its Docker tab, see Swarm and Compose.

The command palette (ctrl+k) of the list also has the server’s journal and Clean up container settings.