PLEG not always triggering actions

@RichardTSchaefer

I have 2 PLEGs, a HVAC and a Lighting PLEG. I think all the errors so far were on the HVAC one. And i only enabled debug on the Lighting one as that was where i noticed it first… Its always like that… obviously have the wrong one… anyway, i enabled debug on both now. As it was quite big already, and with 2 PLEGs now in debug mode, how can i get it to start fresh ? Unplug Vera’s power ?
I will keep monitoring and email you again when the next error occurs. I am quite sure the errors (or missed schedules) were not down to a reboot, that would be too much coincidence in my opinion, but we will see next time it happens.

Not a Reboot … but a Vera Restart. When I am working on things I might do a restart every couple of minutes …

How do i trigger a restart ? Sorry about the silly question i think i should know… But as its not a reload or a power cycle i am not sure

When you hit the “Reload” button … you are restarting the LuaUPnP program that is running on your Vera.

When you have changes and hit “Save” … it saves the changes in the current running version … then immediately restarts the LuaUPnP program.

It is not actually rebooting Vera. Rebooting is a Linux level restart.

So if you were making configuration changes during the time of an “Absolute” schedule … you could miss the schedule.

Ok thanks for that explanation. I was not at home when that schedule did not fire, so probably unlikely that a reload causedit, and there were too many missed ones lately. I have made a few little changes today which needed a save, so log file should be back to a smaller size. Lets wait for any missed schedules or other actions not firing. I still have this error in my schedule, it did work fine yesterday, but Saturdays is still on true. Very annoying.

StudyWD1 Weekly 18:15:00 1,2,3,4 None Weekly 22:00:00 1,2,3,4 None false 2013-11-14 18:15:00
StudyWD2 Weekly 18:15:00 5 None Weekly 22:20:00 5 None false 2013-11-15 18:15:00
StudyWE1 Weekly 08:00:00 6 None Weekly 22:55:00 6 None true 2013-11-16 08:00:00
StudyWE2 Weekly 13:00:00 7 None Weekly 22:10:00 7 None false 2013-11-17 13:00:00

The size of the log file does not matter to me.
If it’s two big to mail break it into pieces and send me one piece at a time.
I can easily find what I need.

This condition stuck in true (I have another one like that) really screws things up for me, and i do not wait until Saturday when it might rectify itself. Can i make this condition false ? I have tried unticking the day, but it is still true…

StudyWD1 Weekly 18:15:00 1,2,3,4 None Weekly 22:00:00 1,2,3,4 None true 2013-11-18 18:15:00.106
StudyWD2 Weekly 18:15:00 5 None Weekly 22:20:00 5 None false 2013-11-15 18:15:00.100
StudyWE1 Weekly 08:00:00 None Weekly 22:55:00 None true 2013-11-16 08:00:00.100
StudyWE2 Weekly 13:00:00 7 None Weekly 22:10:00 7 None false 2013-11-17 13:00:00.100

LoungeWD1 Weekly 17:45:00 1,2,3,4,5 None Weekly 23:30:00 1,2,3,4,5 None true 2013-11-18 17:45:00.106
LoungeWE1 Weekly 08:00:00 None Weekly 23:30:00 None true 2013-11-16 08:00:00.252
LoungeWE2 Weekly 11:30:00 7 None Weekly 22:45:00 7 None false 2013-11-17 11:30:00.149

LoungeWE1 and StudyWE1 are the culprits… i do not want to delete the condition as that always screws up things when i rewrite it…

Temporarily set the off time to the near future and let it run through that time. Then reset the times to what they should be.

I will add to the Request list to manually override timers.

@Mike123
Have not heard anything lately … do you still think there are problems ?

@RichardTSchaefer

Not since i have the debug enabled on both PLEGs… might just have been a temporary glitch ? Really do not understand it. I will leave the debug on for a little while longer, and do a save once in a while just to keep the log file at reasonable size, but i have a usb stick anyway so should not cause too many problems

I have had 2 problems since this last post, hopefully they have been captured on the log.
The HeatWDOn schedule did not turn off, as a result when HeatWEOn turned true the heating did not turn on
The StudyWD2 schedule did not turn off, so when StudyWE1 turned true the study rad did not turn on.

I will email you the status report and the log file (you might have to look in your junk mail again)

@Mike123
The log file starts at 8:52 AM on 11/23/13
So it missed the events.
If you can check to see if you have some older logs.

Maybe something like: LuaUPnP.log.1.gz (Maybe more like .2.gz .3.gz …)

Since you said you have USB logging … turn of the setting (Setup → Logs) Archive old logs on server.
I think MCV is the only one that can see those log files.

turn of the setting (Setup -> Logs) Archive old logs on server

I have done that now.
But where would i find these files

LuaUPnP.log.1.gz (Maybe more like .2.gz .3.gz …)
, there are not showing where the normal log file is. I have done a file search too, but nothing. I guess they are on the server, and i probably cannot get to them ?

They will now be stored in the Logs area … It will keep the last 10 log files. The .gz extension is a compressed version to save space. Just send these files to me when you see the next problem.
LuaUPnP.log is the current log.
LuaUPnP.log.1.gz is the previous one
LuaUPnP.log.2.gz is the one before that …

The date on the log files is for the LAST entry in the file.
So you should be able to determine which log files to send me.

I think I have got a problem which I finally manage to capture on a log file. An action is supposed to switch on Multiswitch 1 when

StudyHeatOn1 (StudyWD1 or StudyWD2 or StudyWE1 or StudyWE2) 2013-12-04 18:15:00.531 true
StudyHeatOn (StudyWD1 or StudyWD2 or StudyWE1 or StudyWE2) and StudyOn 2013-12-04 18:15:00.917 false

either of these are true. Should really only be StudyHeatOn1 but i forgot to delete the action from StudyHeatOn, but it should not make a difference as the action is to set the multiswitch to 1 (instead of toggle -1)
But at 18:15 the condition became true, but the multi switch was not turned on. Maybe it is me not seeing something obvious, maybe it is a bug, i am not sure. It is quite possible its me not seeing something obvious…

I have emailed you the full status report and log files (PLEG was still on debug)

I have just had something else on my other PLEG, but unfortunately i had debug off on that one, so can only post the status (which probably doesnt really help trying to find the problem…
xmaslights should have turned false last night at 23:00, and turned back on again this afternoon. It must have stayed true all the time, as the timestamp is from yesterday afternoon

Schedules
Name Type Time Days Random On Delay Off After Type Off Time Off Days Random Off Delay State Last Schedule
xmas Weekly +00:45:00t 1,2,3,4,5,6,7 None Weekly 23:00:00 1,2,3,4,5,6,7 None true 2013-12-05 16:34:21.100

xmaslights xmas 2013-12-04 16:34:47.192 true
xmaslightsoff NOT xmas 2013-12-03 23:00:00.193 false

@RichardTSchaefer
I have caught a few more schedules not working as intended, i have sent you a couple of emails (pls check your junk)

I have updated to PLEG 5.5, and used the new ‘do it now’ in schedules to reset the schedules which should have been false but were true. Now all my schedules are as they should. Lets see if that update cures the restarts which probably caused the schedules not to fire properly.
Great new feature. Really helps when you have schedules which for one or another reason have not done what they should have, or to test. Very useful.

Still having problems, this time with my other PLEG.

Although Mikehome and Justhome were both false, autoaway was true, but did not switch the multi switch.

Mikehome ()Mike is present 2013-12-16 18:32:39 false
Justhome (
)Justine?s iPhone is present 2013-12-17 06:13:18 false

AutoAway (NOT Mikehome) and (NOT Justhome) 2013-12-17 10:28:25 true
AutoAway5m AutoAway and (AutoAway; NOW > 5:00) 2013-12-17 10:33:39 true

Actions for Condition: AutoAway5m
Immediate
Device Action Arguments
MultiSwitch 2 SetStatus5 newStatus5=1
MultiSwitch 2 SetStatus6 newStatus6=0
Lighting PLEG RunScene SceneNameOrNumber=84
MultiSwitch SetStatus1 newStatus1=1

I do not know if the scene 84 has been triggered, but none of the multi switches have.

@RichardTSchaefer
I will post a log file to you later, i will send a pm one i have, as its probably going to end up in your junk again
I have not had problems with this PLEG for a while, so that must have something to do with the restarts or errors you have seen