Huntarr Stage Management doesn't work, or I don't understand it

[yroyathon]

When I first setup Huntarr (a few months ago?), I wasn't sure what I was doing exactly so I think I started with state management reset interval set to a low setting like 1 week or 2 weeks, perhaps that was the default or near the default. Sleep time between searches was 20 minutes I believe, perhaps the default. There was a flurry of activity, as there is for everyone first using this app. Once I saw that, I realized I wanted the scope of this search to exist on a longer schedule and to run less often, so I increased the time between searches as well as increased the state management reset interval (first to 1 month, then to 2 months). Currently I am doing 60 minutes of sleep between searches for both Sonarr and Radarr. Some weeks later, I thought Huntarr would get to a quiet period, where it had already run through all of my items in Sonarr and Radarr. Then there'd be no futher Huntarr searches until the state management period ended. But I keep seeing the number of Active Tracked items in the state management go to 0. At first I thought that the previous reset interval on existing items would have to age out first before the new reset interval would be applied. But then a few weeks went by and I'm still seeing the tracked items go back to 0, starting the search over.

I did a little arithmetic to sanity check. I have 2400 movies, Huntarr searches 4 movies at a time for possible upgrades, with a 60 minute sleep delay between the round of 4 searches. So that's 600 rounds of searches, which take about 600 hours or 25 days. So if I look in the history for the upgrade for a single movie, I should only see 1 upgrade search for said movie in the last 25 days. Instead, I see 4 or 5 upgrade searches. For all movies, I had hoped for 25 days of upgrade movie activity, followed by about 35 days of waiting.

So how does state management work? Because it doesn't work in several of the ways I thought it might.

Or maybe state management works fine, but some event keeps unintentionally resetting the actively tracked items to 0, like whenever Huntarr gets updated and my docker manager grabs and loads the new docker image.

Thanks for all your work on the app, it's great, I just need to understand how to work with the state management.

[yroyathon]

I've been collecting some data, and even though I have the reset interval set to 1440 (60 days), what I am observing is that the reset for radarr and sonarr occurs every 7 days. Not what I set, not what I want.

Has anyone else had this issue? And if you fixed it, how?

[castielwaverly]

Yeah, you're right that it doesn't work.

The app decides when to reset state here:

Huntarr.io/src/primary/state.py

Lines 82 to 84 in c6c49dc

	# Use a much longer default interval (1 week = 168 hours) to prevent frequent resets
	reset_interval = settings_manager.get_advanced_setting("stateful_management_hours", 168)

And if you follow that through, you can see it's getting the reset interval from the general settings and not the app settings:

Huntarr.io/src/primary/settings_manager.py

Lines 563 to 564 in c6c49dc

	general_settings = load_settings("general")
	return general_settings.get(setting_name, default_value)

The general settings have 168 hours baked in as the interval:

Huntarr.io/src/primary/default_configs/general.json

Line 16 in c6c49dc

"stateful_management_hours": 168,

As a user, you can access the settings directly via sqlite3 huntarr/config/huntarr.db. For example, here's reading the value directly:

sqlite> select * from general_settings where setting_key = "stateful_management_hours";

id	setting_key	setting_value	setting_type	created_at	updated_at
59	stateful_management_hours	168	integer	2025-08-18 02:27:28	2025-08-18 02:27:28

The path to your DB and whether or not you have the right SQLite tools installed will depend on your setup. You can learn about updating values here: https://www.sqlitetutorial.net/sqlite-update/. Note that this will be a global setting and apply across all apps.

The real fix is to make a change here

Huntarr.io/src/primary/state.py

Line 84 in c6c49dc

reset_interval = settings_manager.get_advanced_setting("stateful_management_hours", 168)

and update it to something like (different instances under the app would have to be accounted for too):

reset_interval = settings_manager.load_settings(app_type).get("stateful_management_hours", 168)

[yroyathon]

@castielwaverly thanks for confirming what I'd seen, and finding the culprit in the code. Also finding the temporary workaround, which I just did. Hopefully that db entry doesn't get overwritten with future huntarr updates. It is odd that a GUI setting in Huntarr basically does nothing. Maybe it used to be just a label and was not for editing. Lots of work went into this huntarr app in a relatively short period of time.

Hopefully that bug and this issue gets fixed in some future release.

[castielwaverly]

@yroyathon Yeah, I see several places where it uses the default config and writes it to the DB, though it looks like it only happens when it's missing. Not 100% sure though!

Looking at the state of the app in general, there's a mix of global, app, and instance level behavior in this state management flow (e.g. global interval is used, app level reset time is used, instance level media IDs are maintained). Based on comments I see, I think there was a global interval at some point in advanced settings (Settings > Advanced Settings), which was then moved to app settings. Then perhaps there was only support for one instance for a while before that was introduced. At each step, if the change was 90% implemented, we would now have code from all three "eras", so to speak.

Anyway, I decided to give this a couple hours and mess around with my own fix for this, which was a bit more involved than my code change suggestion above would suggest 😆. I'll keep poking at it when I get time.