Pausing/unpausing and suppressing certain notifications on the fly

[macdis]

I've been using Dolce for a year or more now. It's great. However, there's one thing I can't seem to get it to do.

I have it running 24/7 but when I run a certain maintenance task, scripted in bash and manipulating other containers, I have to pause/unpause or stop/unstop these other containers. This maintenance task is short, but runs hourly, and I don't want Dolce to notify me about the containers starting, stopping, becoming healthy etc.

What I would like to do is disable Dolce briefly, run the maintenance task, and then re-enable Dolce without getting any messages like:

10:25:20 PM                         
 health_status: healthy container_a
10:25:20 PM                         
 health_status: healthy container_b
10:25:20 PM                         
 health_status: healthy container_c

Here is what I have tried in order to disable/re-enable Dolce briefly:

DOLCE_BLACKOUT_WINDOWS='00:00-23:59' DOLCE_EVENTS='die' docker compose -f "$main_compose_file" up -d dolce 2>&1
docker compose -f "$main_compose_file" pause container_a container_b container_c 2>&1
[do some stuff]
docker compose -f "$main_compose_file" unpause container_a container_b container_c 2>&1
DOLCE_BLACKOUT_WINDOWS='' DOLCE_EVENTS='' docker compose -f "$main_compose_file" up -d dolce 2>&1

This does not work as intended. Dolce still sends the above health_status messages when it comes back up.

I should mention that I don't get messages about Dolce itself because I run it with

    labels:
      - dolce.disabled=true

In my compose.yaml I have also set the following so that the environment variables are available within the container.

    environment:
      - DOLCE_EVENTS=${DOLCE_EVENTS:-start,die,kill,oom,stop,pause,unpause,health_status}
      - DOLCE_BLACKOUT_WINDOWS=${DOLCE_BLACKOUT_WINDOWS:-}

A couple of notes:

1. DOLCE_EVENTS=${DOLCE_EVENTS:-start,die,kill,oom,stop,pause,unpause,health_status} doesn't refer to anything in my .env file. I just included it so the variable would be available in the container, and so that the default options would apply in post-maintenance circumstances.
2. I put DOLCE_EVENTS='die' in the docker compose command simply so that the environment variable wouldn't be empty, which would apply the default options as per my previous point.

The result, in any case, is that I can't seem to get Dolce to suppress the health_status messages mentioned above. I don't want to disable them globally, just in this one case.

Is there a way to do this?

[dangrie158]

I think the problem is that you have the BLACKOUT_WINDOW set for the whole day. This way dolce never knows the "normal" state for your containers and notifies you that the container state changed during the blackout window (from unknown to healthy).

I guess I should update the last known state in that case, however a fix for you should be to just define windows to only block out events in the timeframe you actually want to ignore, e.g.

DOLCE_BLACKOUT_WINDOWS =00:59-01:01,01:59-02:01.02:59-03:01,....

Maybe seems a bit silly but your use case of hourly maintenance tasks probably requires this anyway :)

Let me know if this fixes the issue you are seeing

[macdis]

Yes, that works! It'd be great if there were a way to update the last known state in such use cases, as you say, but this works for now! Thanks!

[macdis]

Actually, the proposed workaround does not work consistently. For example, I have xx:20-xx:27 as the blackout window, and the maintenance script runs at xx:21 and ends, for example, at xx:26:01. At that moment, Dolce still sends me health_status updates, even though I have omitted health_status in DOLCE_EVENTS. Here's my command to change monitoring to ignore health_status for a few minutes:

DOLCE_BLACKOUT_WINDOWS='00:20-00:27,01:20-01:27,02:20-02:27,03:20-03:27,04:20-04:27,05:20-05:27,06:20-06:27,07:20-07:27,08:20-08:27,09:20-09:27,10:20-10:27,11:20-11:27,12:20-12:27,13:20-13:27,14:20-14:27,15:20-15:27,16:20-16:27,17:20-17:27,18:20-18:27,19:20-19:27,20:20-20:27,21:20-21:27,22:20-22:27,23:20-23:27' DOLCE_EVENTS='start,die,kill,oom,stop,pause,unpause' docker compose -f "$main_compose_file" up -d dolce 2>&1

(Notice that health_status is omitted in DOLCE_EVENTS=.)

I then get a notification like this (this one is time-stamped 3:26 pm, so within the blackout window):

**Dolce Container Status Summary**

> 3:26:13 PM                         
>  health_status: healthy container_a
> 3:26:13 PM                         
>  health_status: healthy container_b

Next updates not before 1/17/2026, 3:26:43 PM.

Any idea what to try next? Thanks!