Reactor resets at midnight and all time conditions revert

So I have a reactor that senses a light level and then closes a bunch of blinds using the >= condition for the lumens

Everything is fine until midnite when precisely at midnight the blinds all open.

There is no “false” condition. It simply is a one way fire. In the morning, there is another reactor that opens only half the blinds that were closed, at a certain time.

There is no logic to undo the total initial close.

Any ideas?

I can add an additional time condition for generic" is it day or night" but that would seen rather redundant to keep reactor from firing at midnight.

When asking a question such as this, you should post a Logic Summary (link in Troubleshooting section of Tools tab).

As a point of comparison with your blinds logic, I also use Reactor to control our blinds and it works flawlessly. Below is an example of the east-facing blinds in a tri-window. They close after sunset, open before pre-dawn golden hour so we can see the sunrise over the ocean. The blinds also remain open if we are entertaining because the palms outside the window are up-lit (also controlled by Reactor) that adds some nice ambiance. Entertaining is a virtual switch that is turned off when we go to bed (so, set before friends arrive, and shut-off after we clean up and go to bed) so the blinds close.

Window Coverings Reactor (Conditions - excerpt)
WindowCoveringReactorConditions
Window Coverings Reactor (Actions - excerpt)
WindowCoveringReactorActions

sorry rigpapa. was in church pondering.

just found the problem. a legacy scene appears to have been called. let me work this out and i’ll post an update

OK. I would also read this: Type: Date-Time - Reactor

Without seeing your report, if I had to guess, I’d bet you’re using an after operator on your timing test(s), and maybe a before as well? Be aware that the context for date-time tests that specify only hours and minutes is within the current day, so midnight is an implicit boundary for those tests.

Here’s the close trigger:

*************************************************** REACTOR LOGIC SUMMARY REPORT ***************************************************
   Version: 3.3hotfix-19217 config 19178 cdata 19082 ui 19195 pluginDevice 6
    System: openLuup version 190213 on Linux veraplus 4.4.0-171-generic #200-Ubuntu SMP Tue Dec 3 11:04:55 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux; loadtime 1550368506; systemReady 1583112916; ALTUI v2.45; Lua 5.1
Local time: 2020-03-01T20:39:40-0500; DST=0; Greenwich, England UNITED KINGDOM
House mode: plugin 1; system 1; tracking off
  Sun data: {"stamp":2020061,"civdawn":1583043072,"nautdawn":1583040758,"sunset":1583084412,"nautdusk":1583088738,"latitude":51.48,"astrodusk":1583091069,"longitude":0,"civdusk":1583086424,"astrodawn":1583038426,"sunrise":1583045083}
  Geofence: not running
     Power: ?, battery level ?
====================================================================================================================================
Light level trigger close (#24) armed
    Version 19082.23 02/27/20 22:31:32
    Message/status: Not tripped
    Condition group "Activate West Side" (AND) false as of 18:27:26 <root>
      &-F-service Light Sensor (10298) urn:micasaverde-com:serviceId:LightSensor1/CurrentLevel > 3500 [72 => 0 at 19:27:25; F/F as of 18:27:26/18:27:26] <condols5pkv>
    Activity root.true
        Device MasterBlackouts (10257) action urn:upnp-org:serviceId:WindowCovering1/Up(  )
        Device MasterDoor (10259) action urn:upnp-org:serviceId:WindowCovering1/Up(  )
        Device FishbowlBlinds (10279) action urn:upnp-org:serviceId:WindowCovering1/Up(  )
        Device GRBlinds (10278) action urn:upnp-org:serviceId:WindowCovering1/Up(  )
        Device Mediaroom (10280) action urn:upnp-org:serviceId:WindowCovering1/Up(  )
    Events
        03/01/20 20:35:10 reload: notice=Luup reload
        03/01/20 20:35:10 start: 
    Devices
        MBRsensor (10296) urn:schemas-micasaverde-com:device:MotionSensor:1 (4/3); parent 1; plugin -
        Light Sensor (10298) urn:schemas-micasaverde-com:device:LightSensor:1 (18/0); parent 10296; plugin -

Here’s the open trigger:

*************************************************** REACTOR LOGIC SUMMARY REPORT ***************************************************
   Version: 3.3hotfix-19217 config 19178 cdata 19082 ui 19195 pluginDevice 6
    System: openLuup version 190213 on Linux veraplus 4.4.0-171-generic #200-Ubuntu SMP Tue Dec 3 11:04:55 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux; loadtime 1550368506; systemReady 1583112916; ALTUI v2.45; Lua 5.1
Local time: 2020-03-01T20:40:58-0500; DST=0; Greenwich, England UNITED KINGDOM
House mode: plugin 1; system 1; tracking off
  Sun data: {"stamp":2020061,"civdawn":1583043072,"nautdawn":1583040758,"sunset":1583084412,"nautdusk":1583088738,"latitude":51.48,"astrodusk":1583091069,"longitude":0,"civdusk":1583086424,"astrodawn":1583038426,"sunrise":1583045083}
  Geofence: not running
     Power: ?, battery level ?
====================================================================================================================================
Open blinds (#11) armed
    Version 19082.19 03/01/20 09:40:27
    Message/status: Not tripped
    Condition group "MBRBlackouts" (AND) TRUE as of 01:45:00 <root>
      &-T-sun after sunrise+20, [1583112911 => 1583112930 at 20:35:30; T/T as of 01:45:00/01:45:00] <condolub5bf>
    Activity root.true
        Device FishbowlBlinds (10279) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
        Device Mediaroom (10280) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
        Device GRBlinds (10278) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
    Events
        03/01/20 20:35:10 reload: notice=Luup reload
        03/01/20 20:35:10 start: 
        03/01/20 20:35:29 action: action=Restart
        03/01/20 20:35:29 start: 
        03/01/20 20:35:31 action: action=Reset
        03/01/20 20:35:31 sensorstate: state=false

Let’s see if it works during the next 2-3cycles.

Thanks rigpapa. I was aware of the daily midnight boundary. Since there is no false directive for each, I would expect them to reset and hold until met in the next cycle…

-d

Your local timezone appears to be set to GMT-5, but your location is longitude 0. Can you confirm those settings? The sunrise/sunset computation relies on both correct timezone and lat/lon location.

will check. they went up at midnight again.

After you get the lat/long location settings right, you’ll need to clear out the sundata variable on the Reactor master device (just set it to “0”) or it won’t update the internal parameters until the next day.

Whenever a ReactorSensor does not behave as expected, go immediately to the Tools tab and run a Logic Summary and review the “Events” section. It will tell what has changed and what actions it took. This should lead you to the issue. Do not restart Luup or the ReactorSensor before running the report, or you will destroy that history.

did as you said. fixed the midnight thing. but now, the up won’t trigger although the condition is true.

two things I noticed: 1) the condition read true to go up, twenty minutes earlier, but did not trigger until I forced a luup reset.

  1. for 4 days, I have gotten the notice to upgrade AltUI, I accept it, and it doesn’t update. Not sure if that has something to do with it all.

Immediately upon luup restart, it triggered and went up immediately.

Last night I realized that the checkbox for test was still checked (this am at 01:00) and reactor was triggered from 2 days ago. fixed that and reset the reactor (hence the log below).

Condition went true at 06:55 but did not open until restart.

I added the between times because I thought the sensor was resetting (because of the checkbox I left on the test case); figured that out but left it in place. I don’t think its needed,and likely will remove.

*************************************************** REACTOR LOGIC SUMMARY REPORT ***************************************************
   Version: 3.3hotfix-19217 config 19178 cdata 19082 ui 19195 pluginDevice 6
    System: openLuup version 190213 on Linux veraplus 4.4.0-174-generic #204-Ubuntu SMP Wed Jan 29 06:41:01 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux; loadtime 1550368506; systemReady 1583302550; ALTUI v2.45; Lua 5.1
Local time: 2020-03-04T01:18:16-0500; DST=0; Greenwich, England UNITED KINGDOM
House mode: plugin 1; system 1; tracking off
  Sun data: {"stamp":2020062,"civdawn":1583129344,"nautdawn":1583127031,"sunset":1583170918,"nautdusk":1583175240,"latitude":51.48,"astrodusk":1583177575,"longitude":0,"civdusk":1583172926,"astrodawn":1583124696,"sunrise":1583131353}
  Geofence: not running
     Power: ?, battery level ?
====================================================================================================================================
Open blinds (#11) armed tripped
    Version 19082.20 03/04/20 01:17:13
    Message/status: Tripped
    Test time set: 2020-03-02 06:55
    Condition group "MBRBlackouts" (AND) TRUE as of 03-02.02:02:00 <root>
      &-T-sun after sunrise+0, [1583143056 => 1583150100 at 03-02.06:55:00; T/T as of 03-02.02:02:00/03-02.02:02:00] <condolub5bf>
    Activity root.true
        Device FishbowlBlinds (10279) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
        Device Mediaroom (10280) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
        Device GRBlinds (10278) action urn:upnp-org:serviceId:WindowCovering1/Down(  )
    Events
        03/04/20 01:15:44 reload: notice=Luup reload
        03/04/20 01:15:44 start: 
        03/04/20 01:16:32 action: action=Reset
        03/04/20 01:16:32 sensorstate: state=false
        03/04/20 01:17:14 configchange: 
        03/04/20 01:17:15 sensorstate: state=true
        03/04/20 01:17:15 startscene: scene=root.true, sceneName=root.true
        03/04/20 01:17:15 runscene: scene=root.true, sceneName=root.true, group=1, notice=Starting scene group 1
        03/04/20 01:17:15 endscene: scene=root.true, sceneName=root.true

It seems test time is still on. I also would encourage you to upgrade to the current 3.5 release.