This has been confirmed as a bug in firmware 1.6.641 and has been forwarded to the software developers for a future fix.
“It seems there is an issue in UI6 which causes LuaUPnP to crash when a Lua scene is active with a recurrent schedule (e.g. every X minutes).”
Wrote a very simple script that compares the temperature on the first floor to that of the second floor and kicks on some low powered in-line duct fans if the second floor is hotter than the first. Works great to balance temperatures on both floors.
The problem I’m having is that if I schedule this script to run every 15 minutes it causes LUUP (and the Vera) to lock up, but only after it’s been running for many hours. I’ve been doing a Vera Reset and Restore from backup to get it back to functioning.
Anyone know why this would lock LUUP up when put on a 15 minute schedule?
local tup = luup.variable_get("urn:upnp-org:serviceId:TemperatureSensor1","CurrentTemperature", 58)
local tdn = luup.variable_get("urn:upnp-org:serviceId:TemperatureSensor1","CurrentTemperature", 73)
tup = tonumber(tup)
tdn = tonumber(tdn)
if (tup > tdn) then
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“1” },76)
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“1” },77)
else
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“0” },76)
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“0” },77)
end
return true
Device 58 is an Everspring ST814 Temperature/Humidity Detector.
Device 73 is an EcoBee Thermostat connected through the plug-in.
Devices 76 and 77 are Aeon Labs DSC06106-ZWUS Smart Energy Switches.
The fact it takes hours to happen makes me think it’s a resource issue.
[quote=“AEBogdan, post:1, topic:181576”]Wrote a very simple script that compares the temperature on the first floor to that of the second floor and kicks on some low powered in-line duct fans if the second floor is hotter than the first. Works great to balance temperatures on both floors.
The problem I’m having is that if I schedule this script to run every 15 minutes it causes LUUP (and the Vera) to lock up, but only after it’s been running for many hours. I’ve been doing a Vera Reset and Restore from backup to get it back to functioning.
Anyone know why this would lock LUUP up when put on a 15 minute schedule?
local tup = luup.variable_get("urn:upnp-org:serviceId:TemperatureSensor1","CurrentTemperature", 58)
local tdn = luup.variable_get("urn:upnp-org:serviceId:TemperatureSensor1","CurrentTemperature", 73)
tup = tonumber(tup)
tdn = tonumber(tdn)
if (tup > tdn) then
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“1” },76)
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“1” },77)
else
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“0” },76)
luup.call_action(“urn:upnp-org:serviceId:SwitchPower1”,“SetTarget”,{ newTargetValue=“0” },77)
end
return true
Device 58 is an Everspring ST814 Temperature/Humidity Detector.
Device 73 is an EcoBee Thermostat connected through the plug-in.
Devices 76 and 77 are Aeon Labs DSC06106-ZWUS Smart Energy Switches.
The fact it takes hours to happen makes me think it’s a resource issue.[/quote]
have you been able to capture the vera log on this restart?
It may be resource limitations but from something else
I unplugged and replugged the two Aeon Labs DSC06106-ZWUS Smart Energy Switches, the ones controlled by the script above, and now I no longer see the memory leak error in the logs.
*** UPDATED ***
NOPE. Memory leak errors returned after a while…
No help from support.
Been trying to figure this out for a week now. May need to slowly rip my Vera setup apart piece by piece or reset the VERA and start from scratch.
Your script only retrieves the temperature from the Ecobee thermostat but does not otherwise exercise the code in any particular way. My Vera, with one Smart SI thermostat and a standard array of switches, cameras, motion sensors, a Nest, etc., does not show any stability problems. It’s not to say that there isn’t any odd interaction behind the scenes with the plugin, but I tend to think the problem lies elsewhere.
I see the lines in your logs:
01 06/13/14 10:41:39.222 luup_log:66: ecobee: Error: nil: no valid JSON value (reached the end) <0x2d121680>
06 06/13/14 10:41:39.223 Device_Variable::m_szValue_set device: 66 service: urn:ecobee-com:serviceId:Ecobee1 variable: status was: 1 now: 0 #hooks: 0 upnp: 0 v:(nil)/NONE duplicate:0 <0x2d121680>
01 06/13/14 10:41:39.224 luup_log:66: ecobee: Unable to getThermostatSummary; skipping status update. <0x2d121680>
But I think this is likely due to an out-of-memory situation caused by some other leaking code. You might have to turn off the plugin Richard pointed out to see if that’s the culprit.
The logs seem to point toward standard physical switches as causing the issue. I’m now getting messages like these fairly regularly from various devices.
I’ve only owned the Vera for about a month. Had trouble when I tried to revert it back to version 5, so I left it on version 1.6.641.
Have the following plug-ins:
Global Cache GC100 1.0
Wake-on-LAN 1.62
Yamaha Network Control Alias Plugin 2.5
Foscam IP Camera 2.6
VistaCam SD, PT & HD for UI6 1.82
VistaCam SD, PT & HD for UI5 1.44
VeraAlerts 5.7
Squeezebox Control 1.0
MultiSwitch 1.3
Countdown timer 6.0
ecobee Thermostat 0.9
MiOS Update Utility
[quote=“AEBogdan, post:8, topic:181576”]The logs seem to point toward standard physical switches as causing the issue. I’m now getting messages like these fairly regularly from various devices.
I’ve only owned the Vera for about a month. Had trouble when I tried to revert it back to version 5, so I left it on version 1.6.641.
Have the following plug-ins:
Global Cache GC100 1.0
Wake-on-LAN 1.62
Yamaha Network Control Alias Plugin 2.5
Foscam IP Camera 2.6
VistaCam SD, PT & HD for UI6 1.82
VistaCam SD, PT & HD for UI5 1.44
VeraAlerts 5.7
Squeezebox Control 1.0
MultiSwitch 1.3
Countdown timer 6.0
ecobee Thermostat 0.9
MiOS Update Utility[/quote]
you may want to put more effort to getting back to the mores stable version 5. Having tried the upgrade, I was frustrated and reverted too. Not ready for prime time that version 6, in my opinion.
plus, you may want to see where you are with system resources, particularly if you are on a vera lite.
It appears this did have something to do with device 65, the Aeon Labs Aeotec Z-Wave Micro Smart Energy Switch. I removed that switch and all the memory leak errors vanished from the logs.
Wonder if the device was corrupting the energy monitoring process as a whole, making all energy monitoring switches unstable, including the two turned on and off by the Scene above.
Sure learned a lot about: Resets, Restores, Healing, monitoring the Logs and System Memory
After further testing, I have come to the conclusion that this issue is caused by any Aeon Labs Smart Energy Switch, like the ones that are turned on by the script listed above. I can create Memory Leak errors in the logs by turning any of my Smart Energy switches on and off.
This would explain why running the script would causes my Vera to become unstable. I am using the current latest version of the Vera firmware, 1.6.641, and have the MiOS Update plug-in installed, so this should be under the latest Vera code.
Looks like the memory leaks were not the problem, as LUUP locked up even after removing the switch.
I decided to try a factory refresh and rebuild my system from scratch. As soon as I added the above scene and assigned it to a 15 minute schedule LUUP started locking up again.
Decided to try an experiment. I replaced the 15 minute schedule with a Countdown Timer plugin assigned to a 15 minute duration that repeated. Vera has been running fine for a few days now. I’ve been able to edit scenes again without LUUP freezing my Vera.
*** UPDATE 6/24/2014 *****
This has been confirmed as a bug in firmware 1.6.641 and has been forwarded to the software developers for a future fix.
“It seems there is an issue in UI6 which causes LuaUPnP to crash when a Lua scene is active with a recurrent schedule (e.g. every X minutes).”
I also experienced this issue and have been fighting it for the last 2 days. Thank you very much for your post! Hopefully this gets fixed in the next release.
Best Home Automation shopping experience. Shop at Ezlo!