You would need to evaluate about 20,000 conditions between all of your PLEG/PLTS modules that reference NOW before something like NOW could cause an impact on Vera … like causing it to miss a schedule and restart.
It takes about 1/2 ms (or less) to evaluate a condition. More time if it fires actions as a result of evaluating the conditions.
1/2 ms is for a very complicated expression (Lots of terms!)
So if you have 30 conditions … it takes less than 15 ms to evaluate.
The number of conditions or the use of NOW is not going to significantly effect the stability of your system.
The number of Actions fired/second or the amount of free memory will effect the stability of your system!
Excellent, thanks RTS! I will wipe that from my thought process then. I will how ever look how many actions are being triggered as I do have some _Name conditions as my trust in some of these devices i have is low (Boiler Actuator and Danfoss TRV’s). I will monitor their actions.
Thanks Rex, I will tie that a go. That is a damn good idea. I keep forgetting abut “Not”!
OK I have my mem down to about 80% now for the LuaUPnP process by removing loads of stuff (plugins and devices).
With 5.6 of PLEG no more errors in the log (not that it was an issue just something Mios pointed out)
But still my Vera crashes all the time.
Linked is something to do with “Locks” and 'Deadlocks" which i think is when it reboots: Vera Crashing - Pastebin.com
Anyone know what this means or any ideas how to stop it?
There are other causes to reboots … memory was the primary.
The others have to do with to many requests to the same instance of a plugin … where the plugin was not designed to properly handle these.
Plugins that make web requests are candidates for this type of problem.
Examples are Vera Alerts, Ping, Nest, Wifi Lights, Push, Prowl, LAN connected Alarm Systems, …
I would look at those that get called a lot and that may have times where the response to the external device is slow.
I can tell you I have had to make changes for Vera Alerts to reduce the probability that it is the source of the problems.
I do not know if others have done the same.
Can you correlate some plugins with the times of the crashes ?
So I have removed all things like Ping, IOSLocator Plugin and a few others of that type to bring the mem down which it has (that and removing all my cameras and stuff which is mainly cosmetic from when i first started).
All i have left is Virtual Switch, Push Notifications (Prowl), 4 of your plugins (Day and Night, PLEG, PLEG Core and PLTS), DSC Alarm Panel, Twillio and Datamine so I am going to focus on the DSC Plugin next then as I don’t push very much anymore, just that it has rebooted.
So as DSC is very chatty as its all my PIR’s and Door sensors, I will focus on that as it is connected to a EoP device due to its location so i wonder if the Plugin has issues talking to the alarm panel. I know the DSC panel reboots now and then and i have a replacement board as i can’t find the cause of the reboots. Ill change that and add some notifications if i can for when it reboots .
Still waiting for Mios to come back too so hope they find something to help pin point the issue.
I think I’ve had these reboots since day one with vera, it’s just taken me a while to come to this conclusion.