Tutorial: Single scene instance / Event model

[Sjakie 156]

hI Jan,

If I update  Supervisor B

Event Runner B52

It stops all scenes and put them on manually, restarts every 20 seconds.

Back on previous version and all runs smoothly.

[Sjakie 156]

Hello Jan,

I update Supervisor >>B3

Event Runner >>>>>>>b52

Result is all scenes restarts every 20 seconds orso and stops and will be set on manual.

Restored back in Supervisor B2

Event Runner B46 and all runs smoothly again

[FrankT 0]

@jgab when running my test program in ZBS (not in speedmode) there seems to be a difference between the PC clock  and the internal clock in the simulation. 

Please login or register to see this attachment.

[jgab 2,642]

Yes, the simulated clock will drift quite a lot (I think of it as almost realtime ) . It's a tricky problem but I will look into it again and see if it can be done better.

[jgab 2,642]

15 hours ago, Sjakie said:

Hello Jan,

I update Supervisor >>B3

Event Runner >>>>>>>b52

Result is all scenes restarts every 20 seconds orso and stops and will be set on manual.

Restored back in Supervisor B2

Event Runner B46 and all runs smoothly again

 

This is strange because the ping/keepalive mechanism was not changed.

However, try to update to Supervisor B4 that I just uploaded. If you have scenes that are very slow to respond to the Supervisor you can try to increase the timeout, it's now set to 10s.

Please login or register to see this code.

 

[jgab 2,642]

6 hours ago, jgab said:

Yes, the simulated clock will drift quite a lot (I think of it as almost realtime ) . It's a tricky problem but I will look into it again and see if it can be done better.

 

Ok, I have pushed a new version of HC2.lua (fix5) that has a much beter realtime clock. I run it for a few minutes and no drift. The old one drifted a couple of seconds per minute...

However, if you stop at a breakpoint or do other things that interrupt the emulator (like clicking on web GUI buttons) it will  drift (emulated time freezes) because we want the time to be consistent from the emulated scene's perspective.

[FrankT 0]

8 hours ago, jgab said:

 

Ok, I have pushed a new version of HC2.lua (fix5) that has a much beter realtime clock. I run it for a few minutes and no drift. The old one drifted a couple of seconds per minute...

However, if you stop at a breakpoint or do other things that interrupt the emulator (like clicking on web GUI buttons) it will  drift (emulated time freezes) because we want the time to be consistent from the emulated scene's perspective.

Ok thank you. Will try next weekend, away on a trip this weekend.

[Sjakie 156]

Hello Jan,

Sorry for the two posts, I am away from home and the internet here is not that what you expect.

Supervisor B4 gives me this error:

Please login or register to see this image.

/monthly_2019_05/image.png.79d1ebe49e013d4b745b3b90b65f8e0f.png" alt="image.png.79d1ebe49e013d4b745b3b90b65f8e0f.png" />

[jgab 2,642]

45 minutes ago, Sjakie said:

Hello Jan,

Sorry for the two posts, I am away from home and the internet here is not that what you expect.

Supervisor B4 gives me this error:

Please login or register to see this link.

 

You manage to produce the darnest error  

This is in principle an impossible error that happened anyway... I will have a look at it tonight and see if I can reproduce it. 

Does your HC2 box have a heavy CPU load?

[Sjakie 156]

Hello Jan,

I am not surprised!!!

When I changed RaspberryPi into Raspian it started also with install of -npm. I have the idea Murphy is with me.

Prove is I mentioned to my daughter buy hue lamps piece of cake to install. Okay dad buy it and let it shine my house.

It toke me 3 hours included 1 hour with the helpdesk to let it work>>hhehe

Supervisor B2 was running without any problem or notification.

The cpu load is lower as before I started 70-100%. According Zwave monitor:

Please login or register to see this image.

/monthly_2019_05/image.png.fd88e8acef86384c0904c12873e19a4a.png" alt="image.png.fd88e8acef86384c0904c12873e19a4a.png" />

 

It will go also sometimes to 100%

27% used 14% cached 2% buffer 57 % free

6 ER scenes , 1 supervisor, 2 lua scenes security, homeTable, and 6 GV scenes

I will  wait for your advice

 

 

 

 

 

[jgab 2,642]

20 hours ago, Sjakie said:

Hello Jan,

Sorry for the two posts, I am away from home and the internet here is not that what you expect.

Supervisor B4 gives me this error:

 

As I said, the error at that line in the code is very strange. I'm listing all ER scenes with api.get("/scenes") and then I'm retrieving a specific scene by api.get("/scenes/X") and it returns nil... Anyway, I've pushed a new version of Supervisor with a check so it shouldn't crash - it will also log with red what sceneID made the error. If you see it please note it down and tell me if it is something strange with that particular scene. 

[jompa68 121]

On 5/26/2019 at 8:48 PM, jgab said:

I'm testing this in my main lighting scene at the moment, and it seems to work... to make sure that I'm running my own latest version.

A bit risky, and I will soon move over to a stable and beta model for the code.

Please login or register to see this code.

 

I notice new version is out of ER, but VERSION.json file is not updated so rule is not applying. Forgotten?

[jgab 2,642]

4 minutes ago, jompa68 said:

I notice new version is out of ER, but VERSION.json file is not updated so rule is not applying. Forgotten?

Yes, forgotten - fixed. Thanks. (must find someway to automate that)

 

Edited May 30, 2019 by jgab

[jgab 2,642]

2 minutes ago, jgab said:

Yes, forgotten - fixed. Thanks. (must find someway to automate that)

 

..and I also noted that it also takes some time for me before the checkVersion routine will get the new uploaded version - I just checked now (~3min later) and I still receive the old VERSION.

[jompa68 121]

Just now, jgab said:

..and I also noted that it also takes some time for me before the checkVersion routine will get the new uploaded version - I just checked now (~3min later) and I still receive the old VERSION.

Yeah, same here.

[jgab 2,642]

1 minute ago, jompa68 said:

Yeah, same here.

Do you have a better idea how to solve this? I think I remember that you had a version check in one of your scenes?

Just now, jgab said:

Do you have a better idea how to solve this? I think I remember that you had a version check in one of your scenes?

Ha, now I got the new version. 12min later.

[jompa68 121]

24 minutes ago, jgab said:

Do you have a better idea how to solve this? I think I remember that you had a version check in one of your scenes?

I did have checkversion on my own server. Think GitHub have a bunch of servers and will take a while before all are updated.

Now i got the update. Cross my fingers now that it works 

 

Function Util.patchEventRunner did not update scene, no error

 

Sorry, my mistake

did miss this part in my scene. All ok and working. Great @jgab

Please login or register to see this code.

 

 

Edited May 30, 2019 by jompa68

[jompa68 121]

Another thing in Supervisor.lua 

Please login or register to see this code.

If i want to use this when scene is restarted or for some reason stops after xx retry, where to implement in the code?

[jgab 2,642]

12 minutes ago, jompa68 said:

Another thing in Supervisor.lua 

Please login or register to see this code.

If i want to use this when scene is restarted or for some reason stops after xx retry, where to implement in the code?

 

Ops, that handler have disappeared...

Try to add (will include in next version)

Please login or register to see this code.

scene is a table; {id=<sceneID>, name=<sceneName>, restarts=<times scene have been restarted>}

[jompa68 121]

Working