Welcome to Smart Home Forum by FIBARO
Search the Community
Showing results for tags 'ram'.
228 downloadsThe purpose of this scene is to monitor CPU and RAM performances. In case the performance decreased it will notify the user before the system becomes unresponsive and/or stuck. During introduction of last few releases many users have an issue with CPU overloaded and system stuck, so the idea is to catch and try to solve the problem before the system doesn't respond. The program execution has two options: 1. Stand alone using local json table 2. Using global variable for json table Using second option you can acces to current min, max, average values of CPU and current RAM load from any other sense or virtual device. --================= User configurable parameters ============================ gVarName = "" ---------------------- Global Predefined Variable name. eMail = true -------------------- send eMail notification to admin only pushNote = true -------------------- send push notification to users in push list popupNote = true -------------------- send popup notification to all mobiles sleepTime = 60 ---------------------- time between samples in seconds. batchNum = 2 ----------------------- batches/readings number per sample. batchDelay = 10 ---------------------- time between batches in seconds. ramHi = 75 ---------------------- RAM high limit. highLim1 = 50; highLim2 = 50 ---- CPU 1,2 highest load limit. normLim1 = 20; normLim2 = 20 ---- CPU 1,2 normal load limit. sample2avg1 = 20; sample2avg2 = 20 ---- CPU 1,2 number of samples to average. --============================================================================ How it works?: (based on default values of user configurable parameters) The program checks if any global variable defined, if not local table in use. Every 60 seconds (sleepTime) the program will get two readings (batchNum) within 10 seconds (batchDelay) of CPU 1 and 2 load. After 20 samples (sample2avg1/sample2avg2) average load of each CPU and RAM is calculating and in case the CPU load is over 50% (highLim1/highLim2) and/or RAM is higher than 75% (ramHi) notification will be send out by eMail, push notification and popup mobile notification. Please note that popup notification will be received by all mobiles in list of the system. If after overload the CPU average load drops below 20% (normLim1/normLim2) and/or RAM drops below 75%, user will be notified as well. Change default values according to your needs, but please pay attention if you reduce the sleepTime and number of samples (sample2avg1/sample2avg2) , the scene could catch momentary peak of load which is not necessarily represents system performances.
Może macie jakiś pomysł co może tak zapychać RAM w centrali HCL? U mnie nawet po wyłączeniu wszystkich scen i restarcie nadal wolne miejsce w RAM 2-3%. Czyli to nie sceny? O co chodzi? Nie ma żadnej diagnostyki tego problemu? Walczę z tym od początku ale wraz z nowymi wersjami oprogramowania sytuacja jest coraz gorsza. Efekt jest taki że praktycznie codziennie muszę restartować centralę. Nie jest to przecież normalny sposób pracy. Zrozumiałbym restrat raz na 2-3 tygodnie. Co wy na to ilość urządzeń około 40 iliść scen około 30
Winter has gone and 30 degrees are back in town so time ago I stopped all scenes that control the heating system in the house. We are talking about 16 scenes changed from automatic to manual, but I did not stop the automatic daily restart scene. Immediately after that, the used RAM dropped from 99% to 80%. Now, looking at the RAM usage I can read numbers between 96 and 99% without implementing anything new. Yes it eats RAM, a lot, not known why, not possible to do anything about that. I read that people with FW 4.130 are less happy then 4.120 or 4.110 or xxxxx. Guys nothing changed, it's always the same story. Don't think that a new FW is totally rewritten from zero every time and in the correct way, it is just a bunch of extra coding squeezed inside the other rows of code. Hardware remains the same. Issues remain the same. Sad but true (cit.)
Hi I have recently upgraded to 4.07 and getting 'out of memory' notifications coming up with RAM close to 98-99%. I am trying to nail the problem and failing so far. I restarted the HC2 this morning with RAM after restart being 27%. I have removed any VDs with loops in them. As regard looping scenes I have got following: Bedroom light on at 25% at 6:20 Mon-Fri Hallway light on for 45s when movement detected with countdown timer checking if there is further movement during 45s Simple time of day scene, setting 'Day' at sunrise and 'Night' at sunset. Is there any way to check what is causing it other than enabling scenes one by one?