When a container is configured with the "restart: always" policy, it seems to fail to attempt a second restart after encountering a failure caused by network-related logging issues. Details:

1. A container with "restart:always" is started and runs successfully for some time. Then the network connection is interrupted. The container continues running but cannot send logs to the remote logging service due to the network outage.
2. The system is shut down and then restarted. After rebooting, the container attempts to restart automatically but fails because it cannot connect to the remote logging service (due to the network still being unavailable at that time).
3. Once the network is restored, the container does not seem to attempt to restart again, even after several minutes, despite the restart: always policy being configured. (sudo journalctl -u docker.service --no-pager, only a single failure entry is observed)

Questions:
What is the retry mechanism for restart: always after a failed restart attempt?
Is there a specific interval or timeout before the next retry occurs?
Shouldn’t the container keep attempting to restart until it succeeds, given the restart: always policy?