Did you go into your SL rule and click Done? You need to do that for it to schedule itself correctly.
Ah ok, I will hang tight (my mind is still trying to unravel the sequencing of events⌠this stuff isnât as easy as seems it should be).
Yes, I did. That rule has been working for a month without being touched. Like I said above, I deleted and rebuilt it and it looks good so far. Weâll see what happens when it sets the schedule for tommorow.
Thank you Bruce.
The anti-Action Sunrise event (turnoff) didnât happen at sunrise; no turn-off event was logged for the lamp. Nothing appears to be scheduled for the next sunrise either. Previous to the update I was seeing the both doAntiAction scheduled along with the doAction. Is this feature working for anyone else?
Sunset events are getting scheduled normally, however; and as of the latest update the day to day changes are being propagated to RM rules correctly as well.
Hereâs what the Simple Lighting rule for âturn off at sunset -6 min with turnoff at sunriseâ looks like this morning (screenshot of the rule is in a prior post; this rule was newly created right after the .705 update was applied on the 17th).
Scheduled Jobs
| Handler | Next Run Time | Prev Run Time | Status | Schedule |
|---|---|---|---|---|
| scheduleSunrise | Tue Mar 20 02:05:29 EDT 2018 | Mon Mar 19 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? |
| doAction | Mon Mar 19 19:00:00 EDT 2018 | PENDING | Once | |
| scheduleSunset | Tue Mar 20 02:05:29 EDT 2018 | Mon Mar 19 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? |
In my case, the antiAction did work this morning and light turned off. My scheduled jobs for the rule look the same as yours.
| Handler | Next Run Time | Prev Run Time | Status | Schedule |
|---|---|---|---|---|
| doAction | Mon Mar 19 19:37:00 EDT 2018 | PENDING | Once | |
| scheduleSunrise | Tue Mar 20 02:05:29 EDT 2018 | Mon Mar 19 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? |
| scheduleSunset | Tue Mar 20 02:05:29 EDT 2018 | Mon Mar 19 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? |
Iâm hoping the doAntiAction gets scheduled at the next 02:05:29 . This might be the way the multiple scheduling events issue was alleviated. I donât know why yourâs didnât work this am.
Well thatâs interesting. Still seems odd that there is no record of the doAntiAction schedule, especially since it appeared before the fix (along with the scores of erroneous scheduled events). Iâll have to dig back into the event log to see if the antiAction fired the first time (Sundayâs sunrise, right after the update). EDIT: no sunrise event happened for the first scheduled sunrise after the rule was installed.
By the way, I notice in a prior post you said you had a âblank doAntiActionâ as expected⌠I never saw this; did this appear in the screen of scheduled events? I take it you since modified the rule to include the antiAction?
Iâve used the AntiAction all along. Yes, when I rebuilt the rule, the blank doAntiAction appeared in the scheduled jobs and I stated âas expectedâ because I assumed it would be scheduled later at 02:05:29.
Ok. Odd that it wasnât there when I created the rule the first time (I created it from scratch after deleting the original; not sure if thatâs what you did when you say you rebuilt it-- perhaps you went in, modified it and re-saved? Might account for the difference there).
So at least at this point neither of us are seeing any doAntiAction on the list of scheduled events.
I deleted it and made a new one. We are seeing the same thing for scheduled jobs but Iâm still very curious as to why your sunrise event didnât happen.
Got it. I misunderstood your post a ways back where you said you didnât have multiple events scheduled; I assume that was referring to my simultaneous RM rule that I had at the time. I no longer have simultaneous events scheduled either, I set my RM rule to fire a minute later than the Simple Lighting rule.
Another thing-- Iâm claiming that my light didnât turn on at sunrise based on there being no event logged for the lamp (I didnât have a logging page open at sunrise and wasnât watching for it). Just to clarify, youâre seeing the turn off event in the device logging or live logging, or both?
Please open the Simple Lighting rule and hit Done. Then show me a screenshot of the app details for Scheduled things.
Iâm seeing it in the device logging. I forgot to leave live logging open last night.
Ok, just did it and voila. There is the doAntiAction I would expect to see. But I canât have created this rule without having hit âDoneâ the first time-- can I?
Scheduled Jobs
| Handler | Next Run Time | Prev Run Time | Status | Schedule |
|---|---|---|---|---|
| scheduleSunrise | Tue Mar 20 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? | |
| doAction | Mon Mar 19 19:00:00 EDT 2018 | PENDING | Once | |
| scheduleSunset | Tue Mar 20 02:05:29 EDT 2018 | PENDING | 29 5 2 * * ? | |
| doAntiAction | Mon Mar 19 07:01:00 EDT 2018 | PENDING | Once |
We are trying to figure out exactly what is going on. Notice that the Sunrise event is there for today. I'm pretty sure that if you check later it will be gone, as it is in the past. Our scheduler should figure that out, and remove it. Then, it would be back to what you saw before. The best thing would be if you could look at this before sunrise tomorrow morning, when it should look like this but with Mar 20 for the actions and Mar 21 for the schedules.
In our testing this works as expected. So I'm not sure why your sunrise event didn't happen, if in fact it didn't happen.
Based on logged device events, it didnât happen, unless there is some kind of command optimization that wonât turn off a light that is already off.
Iâll try to get a capture of what it looks like after 2:05AM or prior to sunrise.
There doesnât appear to be any events logged for the âEventsâ link on the details page for the SL rule itself; Iâm assuming that is normal?
This is an original installation of Simple Lighting (hub #23); I assume that I could delete/re-install it if you advise (I only have 8 SL rules that could be recreated without much trouble).
It is normal for there not to be events logged there. We may change that.
You don't need to reinstall it. It runs off the new code, and I can tell that it is from what you've shown. Not sure you'd get an off event for a device that is off. What driver are you using? And did you have logging enabled for that driver and logs running?
Bingo. As I suspected (in my defense 
I did disclose that the basis for my claim was the device event log and not live logging), device event logging is relevant. Turns out that for two consecutive digital âoffâ events (the source doesnât seem to matter; could be the Device page or a rule, evidently), nothing is logged for second âoffâ.
The live logging, however (at least any level of logging turned on), does record consecutive âoffâ events, and they do indeed get sent to the device-- there is no optimization going on by the driver⌠and thankfully so â otherwise Hubitat would never be able to turn this light off if it was turned on by my Intermatic HA07 Z-Wave remote (paired as a dumb secondary controller to the network) unbeknownst to Hubitat.
So optimization is apparently being done by the Device Event logging (this seems unwise, given there is no other logging being done for the rule), and since the lamp had been turned off prior to the expected sunrise turnoff, it masked its existence (at least Iâm comfortable with that theory until proven otherwise).
Iâm using the Generic Z-Wave Dimmer driver. by the way.
Rule Machine does log events of a rule being run.
