[email protected] Should be modified because of pod creation disabled by default

[yanghl12138]

Describe the bug
Because pod are not created by default#673, [email protected] Should be modified

This is the file generated by podman-compose systemd -a create-unit

[Unit]
Description=%i rootless pod (podman-compose)

[Service]
Type=simple
EnvironmentFile=%h/.config/containers/compose/projects/%i.env
ExecStartPre=-/usr/bin/podman-compose up --no-start
ExecStartPre=/usr/bin/podman pod start pod_%i
ExecStart=/usr/bin/podman-compose wait
ExecStop=/usr/bin/podman pod stop pod_%i

[Install]
WantedBy=default.target

The line

ExecStartPre=-/usr/bin/podman-compose up --no-start

should be

ExecStartPre=-/usr/bin/podman-compose --in-pod=1 up --no-start

To Reproduce

version: "3.1"

services:
  test:
    image: busybox
    command: sh -c "while [[ 1 ]]; do uname -msr; sleep 5; done"

sudo podman-compose systemd -a create-unit
systemctl --user daemon-reload
podman-compose systemd -a register
podman-compose down
systemctl --user start 'podman-compose@busybox'
podman pod stats 'pod_busybox'

Expected behavior
Create pod_busybox and start the service using systemd command.

Actual behavior
No pod created.

Output

> systemctl --user status 'podman-compose@busybox'
× [email protected] - busybox rootless pod (podman-compose)
     Loaded: loaded (/etc/xdg/systemd/user/[email protected]; disabled; preset: disabled)
    Drop-In: /usr/lib/systemd/user/service.d
             └─10-timeout-abort.conf
     Active: failed (Result: exit-code) since Wed 2023-05-17 11:18:51 CST; 4min 15s ago
    Process: 120985 ExecStartPre=/usr/bin/podman-compose up --no-start (code=exited, status=0/SUCCESS)
    Process: 121014 ExecStartPre=/usr/bin/podman pod start pod_busybox (code=exited, status=125)
        CPU: 801ms

May 17 11:18:49 server.fedora podman-compose[120985]: ['podman', 'network', 'exists', 'busybox_default']
May 17 11:18:49 server.fedora podman-compose[120985]: podman create --name=busybox_test_1 --label io.podman.compose.config-hash=e36d575a906>
May 17 11:18:49 server.fedora podman[121007]: 2023-05-17 11:18:49.886335646 +0800 CST m=+0.077186105 image pull beae173ccac6ad749f76713cf44>
May 17 11:18:50 server.fedora podman[121007]: 2023-05-17 11:18:50.183193284 +0800 CST m=+0.374045000 container create 6b3e3ed8dd85f507a09a9>
May 17 11:18:50 server.fedora podman-compose[121007]: 6b3e3ed8dd85f507a09a9f0b1047ea79f3e43ce1d718e5bfd974f238516fb48d
May 17 11:18:50 server.fedora podman-compose[120985]: exit code: 0
May 17 11:18:51 server.fedora podman[121014]: Error: no pod with name or ID pod_busybox found: no such pod
May 17 11:18:51 server.fedora systemd[891]: [email protected]: Control process exited, code=exited, status=125/n/a
May 17 11:18:51 server.fedora systemd[891]: [email protected]: Failed with result 'exit-code'.
May 17 11:18:51 server.fedora systemd[891]: Failed to start [email protected] - busybox rootless pod (podman-compose).

> podman pod stats 'pod_busybox'
Error: unable to get list of pods: no pod with name or ID pod_busybox found: no such pod

Environment:

• OS: Fedora 38
• podman version: 4.5.0
• podman compose version: 1.0.6 (install using dnf)

Additional context

Add any other context about the problem here.

[StaticRocket]

I think the systemd service is fine the way it is, the actual behavior of podman-compose needs to be adjusted such that the pod is created with the containers.

Pods are supposed to track groups of containers, the "project" should map to a single pod, assuming I'm not misinterpreting anything from the docs I've seen.

[mokibit]

@yanghl12138 thanks for opening an issue.
The issue was fixed in #964 by adding a custom section in the compose file so the user can set the preferred value for --in-pod.
If there is no custom --in-pod value in compose file or in command line, --in-pod defaults to True so there is no need to update [email protected] file.
@p12tic the problem can be closed.