Very high MEM usage about iobroker.yahka

[jensweigele]

Can you check if this happens with the current master?
I suspect the logging to be responsible for it (see #53 / #68 )

[captaindork]

Unfortunately still present, to be honest even worse. Any log that would help?

[foxriver76]

I don't think that high RAM usage is caused by the logging. How many devices do you use with YAHKA? Mem usage at adapter start is round about 45 MB growing to 75 MB over time on my system, but I do not use many devices with the adapter. Maybe it somehow correlates with the number of devices.

[captaindork]

Hi, I use ~76 devices in yahka.
and yes, yahka uses around 40-50 MB of RAM during normal use.
My Observartions are rather like that:

• Yahka is started: 35MB
• after ~30s it goes up to 300-500MB for around ~30s
• then it goes back down.

This "peak" in the middle is my issue, because this will lead to crashes if other adapters are started and also eating RAM. So if I start yahka first, wait till it became low again, and "then" start all other adapters, then result is a System which works with 35% free RAM. (Hoping yahka never restarts, because if so it will definetely crash).

While I honestly do not want to mix issues... I am not sure if this is related: Right now since 0.8.1 yahka does not recover from the peak properly at all. it gets restarted, log shows the devices being created and then it crashed within next 30s. Following Log contents (it creates all devices, i leave the last line if this creation for my Vallox Device in, then these errors appear):

2018-12-06 16:41:01.561 - ^[[32minfo^[[39m: yahka.0 adding Vallox 90 SE with UUID: a58baa83-d71a-454a-94da-3a770a858c91 2018-12-06 16:41:25.149 - ^[[31merror^[[39m: Caught by controller[0]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client:url parse http://127.0.0.1:9001 2018-12-06 16:41:25.150 - ^[[31merror^[[39m: Caught by controller[1]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client new io instance for http://127.0.0.1:9001 2018-12-06 16:41:25.151 - ^[[31merror^[[39m: Caught by controller[2]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client:manager readyState closed 2018-12-06 16:41:25.152 - ^[[31merror^[[39m: Caught by controller[3]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client:manager opening http://127.0.0.1:9001 2018-12-06 16:41:25.152 - ^[[31merror^[[39m: Caught by controller[4]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:socket creating transport "polling" 2018-12-06 16:41:25.152 - ^[[31merror^[[39m: Caught by controller[5]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:polling polling 2018-12-06 16:41:25.152 - ^[[31merror^[[39m: Caught by controller[6]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:polling-xhr xhr poll 2018-12-06 16:41:25.153 - ^[[31merror^[[39m: Caught by controller[7]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:polling-xhr xhr open GET: http://127.0.0.1:9001/socket.io/?EIO=3&transport=polling&t=MU4BUX4&b64=1 2018-12-06 16:41:25.153 - ^[[31merror^[[39m: Caught by controller[8]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:polling-xhr xhr data null 2018-12-06 16:41:25.153 - ^[[31merror^[[39m: Caught by controller[9]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:socket setting transport polling 2018-12-06 16:41:25.153 - ^[[31merror^[[39m: Caught by controller[10]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client:manager connect attempt will timeout after 20000 2018-12-06 16:41:25.154 - ^[[31merror^[[39m: Caught by controller[11]: Thu, 06 Dec 2018 15:41:00 GMT socket.io-client:manager readyState opening 2018-12-06 16:41:25.154 - ^[[31merror^[[39m: Caught by controller[12]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:polling polling got data 97:0{"sid":"bWnphHQe2_ErwJVLAACV","upgrades":["websocket"],"pingInterval":25000,"pingTimeout":60000} 2018-12-06 16:41:25.154 - ^[[31merror^[[39m: Caught by controller[13]: Thu, 06 Dec 2018 15:41:00 GMT engine.io-client:socket socket receive: type "open", data "{"sid":"bWnphHQe2_ErwJVLAACV","upgrades":["websocket"],"pingInterval":25000,"pingTimeout":60000}"

This Caught by controller [xy] messages count up to more than 1500 for xy. flooding the log.

[jensweigele]

Please try again with version 0.8.2

[captaindork]

Absolutely Insane @jensweigele !

I an using yahka since the early 0.5.0 releases and can tell: It never ran that good!
Consumption starts with 36, goes up to 110, then 118, then back to 86 and 46. All this within a minute, without errors and without iobroker crashing even if i start it as the last adapter.

Even if yahka always stayed my preferred HK Interface (coming from homebridge-Homematic and also Node-red) this one was really frustrating. You can‘t believe how happy I am -> Thanks for your work on finding this one!!!