Monitor Only doesnt work?

[1mfaasj]

Hello
I used the following YML to launch watchtower, but then NOTHING happens when it needs to do something. Am I missing something?

version: "3.4"
services:
  watchtower:
    image: containrrr/watchtower:latest
    environment:
      - TZ=Europe/Amsterdam
      - WATCHTOWER_MONITOR_ONLY=true
      - WATCHTOWER_NO_PULL=true
      - WATCHTOWER_NOTIFICATIONS=email
      - "[email protected]"
      - "[email protected]"
      - WATCHTOWER_NOTIFICATION_EMAIL_SERVER=smtp.gmail.com
      - WATCHTOWER_NOTIFICATION_EMAIL_SERVER_PORT=587
      - "[email protected]"
      - "WATCHTOWER_NOTIFICATION_EMAIL_SERVER_PASSWORD=mypassword"
      - WATCHTOWER_NOTIFICATION_EMAIL_DELAY=2
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock

the logging:

time="2021-01-26T07:36:56+01:00" level=warning msg="Using `WATCHTOWER_NO_PULL` and `WATCHTOWER_MONITOR_ONLY` simultaneously might lead to no action being taken at all. If this is intentional, you may safely ignore this message."                             
time="2021-01-26T07:37:00+01:00" level=info msg="Starting Watchtower and scheduling first run: 2021-01-27 07:37:00 +0100 CET m=+86403.440599214"                                                                                                                 
                                                 

but nothing happens at the given time ???

[piksel]

Well, that is exactly why that warning message is there. Using both monitor only and no pull tends to not do anything at all.

Since we now have the HEAD requests, it would be possible to have watchtower just do the HEAD digest compare bit here, but that is not something that is currently implemented. If you want to get updates about new images, they have to be pulled.

[1mfaasj]

Thanks for the reply, so when removing that NO PULL line, it should work?

[1mfaasj]

The result is this:
Could not do a head request, falling back to regular pull

What does this mean?

[piksel]

It means that it cannot check if there is a new image by using the a special HEAD request that does not count towards the DockerHub rate limiting quota. If your repository isn't using rate limits it can be disregarded.

[1mfaasj]

I don't understand it completely, does this mean that it is checked whether a new image exists? or is it skipped because of that. I am missing some background information I think.

[piksel]

No, that message by itself doesn't mean it skipped it, just that it used another method of checking. If it failed to check it for some reason there should be a message saying Unable to update container <CONTAINER>: <ERROR>. Proceeding to next.