I think a scene is crashing my HCL?

[dirkies 1]

Hello

this morning I ran a scene (opening shutters on ground floor) and I noticed that the scene took several seconds to start. Then when finished (1 shutter 100% up + 2 shutters 50% up), after about a minute the 2 shutters at 50% open went 100% open by themselves (?)

Right after this, the HCL went completely offline, impossible to access it from android app or webbrowser (WWW just timed out, Android App said "HCL not found"). I saw on the unit itself the internet led on the HCL was off. The network and power led were still on.

I had to power cycle the HCL to make it work again. 

Please login or register to see this image.

/emoticons/default_icon_curve.gif" alt=":-/" />

 

- Last time I had to do this was after I switched on a external lamp with the Android interface, right after that nothing else worked anymore and also here I had to power-cycle the HCL before it responded again.

- Before that it was the Z-Wave that seemingly was dead, the interface and WWW access seemed to work just fine, it just did not transmit any Z-Wave command. The last entry in the log was that it did ran the scene I usually run in the evening aroun 9pm. After that no log entry, and a task supposed to run at 5am did not run anymore. Restarting the services from within the GUI restored Z-Wave traffic.

- The 1st time the unit hung on me was the same day I started using it, it was during scene testing it hung completely. 

Please login or register to see this image.

/emoticons/default_icon_confused.gif" alt=":-?" />

 

Question now is:

since the 2 weeks I have the HCL running at my house, this is now the 4th time the unit crashes. It never stayed OK for more than 5 days before I had to reboot it.

Is there anything that I can do to see why it does this?

If this would be a general issue I am sure others will experience the same thing.

I think this has to do with the scene engine doing something wrong.

 

For now, HCL looks like a nice solution compared with the Vera Edge I had before, but while being a pain to use, the Vera at least never crashed while I used it, and frankly this makes the HCL unusable to me if I have to power cycle it every 4 / 5 days.

Please login or register to see this image.

/emoticons/default_sad.png" alt=":(" srcset="https://forum.fibaro.com/uploads/emoticons/[email protected] 2x" width="20" height="20" />

 

Any suggestions more than welcome!

 

[tom.dewit 0]

First of all, the reboot virtual device also doesn't work when I click on the reboot button in the virtual device. If that is not normal then we even don't have to look at the scene or the variable because it's already malfunctioning in the virtual device itself.

 

My code:

 

POST /api/settings/reboot HTTP/1.10x0D0x0A

Host: 192.168.2.60x0D0x0A

Origin:

Please login or register to see this link.

Content-Length: 00x0D0x0A

Authorization: Basic YWRtaW46YWRtaW4=0x0D0x0A

Referer:

Please login or register to see this link.

X-Requested-With: XMLHttpRequest0x0D0x0A0x0D0x0A

 

I changed the password hash with the hash for admin:admin here instead of my own password.

 

The IP address is 192.168.2.6

 

I also filled in '80' for the port (nothing in the IP address field) --> Look in the attached screenshot

 

Hopefully you can find a mistake here somewhere ...

 

I'm also currently awaiting a call from support for the memory leak problem.

 

Grtz,

Tom

 

 

 

 

 

 

 

Please login or register to see this attachment.

[dirkies 1]

well, mine looks like this:
 
add the IP address in the general tab and remove the linefeeds in your code as they are already there (0x0D and 0x0A). Maybe it acts up as you add twice some CR (in hex 0x0DA) & LF (in hex 0x0A) codes ?

Just make it look as if you want to put the whole code on 1 line (the GUI will automatically wrap text in the textbox as my screenshot shows)

 

my HCL is 192.168.0.39 and my hashed login details are of course different from yours, mine = YWRtaW46U2VjdXJld2F2ZTIwMTUr

 

Make sure there is no character too much at the end of your hashed credentials!

 

Does 

Please login or register to see this link.

 give you a blank screen or an error?

 

and remember I am using a HCL v4.54 but it worked with a v4.41 and v4.53 aswell.

Please login or register to see this attachment.

Please login or register to see this attachment.

[tom.dewit 0]

Hi Dirkies,

 

I really appreciate you helping me out here.

 

In the meantime I had a chat with Fibaro support and they helped me set up the virtual device. You are correct I needed to add my IP address (or 127.0.0.1). However the port needed to be 11111 in my case ... I also needed to remove the extra line breaks.

 

We also found a misbehaving plugin (Onkyo plugin) that was eating away all CPU. I removed it now and let's see if the memory problem returns ...

 

Grtz,

Tom

[s.isaenkov 8]

You can paste even so simple without extra symbols:

Please login or register to see this code.

Don't forget to set Hostname 127.0.0.1 and port 11111 in Advanced tab.

 

And you haven't to add any variables. Just set timer and reboot action in the scene.

[dirkies 1]

why does it need port 11111 while mine works fine on 80 ?

[tom.dewit 0]

I think both will probably work. 11111 looks to me like an internal port that is used (also localhost is used instead of the HCL IP address here)

 

Grtz,

Tom

[mpatel0902 0]

 

@ 

Please login or register to see this link.

 

Does 

Please login or register to see this link.

 give you a blank screen or an error?

 

Yes i put HCL IP

Please login or register to see this link.

and it is blank and i trying your code but not working.

[tom.dewit 0]

I wanted to post a small update after Fibaro support contacted me back in september.

 

It was indeed the Onkyo amplifier plugin that was eating away all CPU. Fibaro support can see that when looking at logs, not accessible by the GUI. The plugin was constantly putting errors in the log. After the plugin was removed, My CPU went down to almost nothing and since september I did not have a single crash anymore. Also all scenes are starting much faster now thanks to the CPU not being pegged.

 

I would really want to make a request to Fibaro engineers to add a log view functionality in the GUI, because this would have been a problem I could have found myself with proper acces to logs.

 

Anyway ... the HCLite is working well now.

 

Thanks dirkies for your help on this matter

 

Grtz,

Tom