Created attachment 91544 [details] energy consumptions stats by apps for 28 hours I think the way how the KDE connect works with PC should be optimized because of too high power consumption. On my Lenovo P780 with KitKat 4.4.2 it spent 22.2% of total energy for 28 hours after the battery has been charged. All time phone was near the PC or the notebook with kdeconnect installed. See screenshot
:( Sorry about that. Did this start happening after KDE Connect 0.8 or was happening with 0.7 as well? Do you know how to reproduce it? Does it always happens or only sometimes? Let's see if we can figure out what's happening... Albert
First thank you for great app! First I get know and setup KDE connect at March 6th from goggle store, so I think it should be 0.8 then it get updated, to 0.8d The power issues I glanced at March 9th. (I suppose it was 0.8d) Stesps to reproduce. I suppose in my conditions t should happens always, I will try to reproduce it and back to you over few days after I will recharge battery on my phone (this is because of power manager app which resets stats after battery get charged) In addition, I tried to see it little more and checked how PC communuctes with the phone and saw 'ping's like below every few seconds.... 16:04:31.650680 IP draken2.softlab.in.ua.47220 > android-a96487215ae2dd7c.softlab.in.ua.1715: Flags [.], ack 418, win 237, options [nop,nop,TS val 53115088 ecr 8951418], length 0 16:04:31.698345 IP android-a96487215ae2dd7c.softlab.in.ua.1715 > draken2.softlab.in.ua.47220: Flags [.], ack 1055, win 529, options [nop,nop,TS val 8951613 ecr 52974393], length 0 16:04:41.698676 IP draken2.softlab.in.ua.47220 > android-a96487215ae2dd7c.softlab.in.ua.1715: Flags [.], ack 418, win 237, options [nop,nop,TS val 53117600 ecr 8951613], length 0 16:04:41.746952 IP android-a96487215ae2dd7c.softlab.in.ua.1715 > draken2.softlab.in.ua.47220: Flags [.], ack 1055, win 529, options [nop,nop,TS val 8951746 ecr 52974393], length 0 16:04:51.762694 IP draken2.softlab.in.ua.47220 > android-a96487215ae2dd7c.softlab.in.ua.1715: Flags [.], ack 418, win 237, options [nop,nop,TS val 53120116 ecr 8951746], length 0 16:04:51.810838 IP android-a96487215ae2dd7c.softlab.in.ua.1715 > draken2.softlab.in.ua.47220: Flags [.], ack 1055, win 529, options [nop,nop,TS val 8951811 ecr 52974393], length 0 16:05:01.826678 IP draken2.softlab.in.ua.47220 > android-a96487215ae2dd7c.softlab.in.ua.1715: Flags [.], ack 418, win 237, options [nop,nop,TS val 53122632 ecr 8951811], length 0 16:05:01.827740 IP android-a96487215ae2dd7c.softlab.in.ua.1715 > draken2.softlab.in.ua.47220: Flags [.], ack 1055, win 529, options [nop,nop,TS val 8951983 ecr 52974393], length 0 16:05:11.842669 IP draken2.softlab.in.ua.47220 > android-a96487215ae2dd7c.softlab.in.ua.1715: Flags [.], ack 418, win 237, options [nop,nop,TS val 53125136 ecr 8951983], length 0 16:05:11.890710 IP android-a96487215ae2dd7c.softlab.in.ua.1715 > draken2.softlab.in.ua.47220: Flags [.], ack 1055, win 529, options [nop,nop,TS val 8952048 ecr 52974393], length 0 for why? why we send zero length packtes? PC & phoe both idle, why we need these pings, why we need to keep connection alive? I will try to look at it little more, will try to see the code.... but I think this will take a wekk or so... Best rehards, Yuri
I have the same issue on 2.3.3. I see that KDEC is responsible for ~25% of the battery usage. I think it started around january/february, but can't be to certain. Before that, the battery usage was negligeable. What I see is that KDEConnect from the phone send every ~0.3s a packet ACK to the PC, which respond with a PSH, ACK. Here is an extract of wireshark output with 0.8h. Note that .12 is the phone and .10 the PC "37","0.297697000","192.168.0.12","192.168.0.10","TCP","66","1714→37388 [ACK] Seq=1 Ack=1 Win=20996 Len=0 TSval=13498 TSecr=390809168" "38","0.297722000","192.168.0.10","192.168.0.12","TCP","484","37388→1714 [PSH, ACK] Seq=1 Ack=1 Win=270 Len=418 TSval=390809244 TSecr=13498" "60","0.611921000","192.168.0.12","192.168.0.10","TCP","66","1714→37388 [ACK] Seq=1 Ack=419 Win=20996 Len=0 TSval=13529 TSecr=390809244" "61","0.611949000","192.168.0.10","192.168.0.12","TCP","484","37388→1714 [PSH, ACK] Seq=419 Ack=1 Win=270 Len=418 TSval=390809323 TSecr=13529" "80","0.958833000","192.168.0.12","192.168.0.10","TCP","66","1714→37388 [ACK] Seq=1 Ack=837 Win=20996 Len=0 TSval=13564 TSecr=390809323" "81","0.958863000","192.168.0.10","192.168.0.12","TCP","902","37388→1714 [PSH, ACK] Seq=837 Ack=1 Win=270 Len=836 TSval=390809409 TSecr=13564" ... I am available if you need more infos.
Just to add a bit. Right now I only have a Keep-alive connection appearing in Wireshark every ten seconds, but the battery still get depleted as fast as before. So a lot of communication isn't the issue. Or maybe it is, by not allowing the wifi to go to stand-by ?
Apparently I cannot observe this bug. Here are the versions of kdeconnect which I currently have in use (I have installed multiple distributions on my notebook): Android: 0.9c from f-droid Debian testing: 0.8-1 Kubuntu 15.04: 0.8-0ubuntu2 Opensuse 13.2: 0.8.99git-34.1 All these 3 Linux setups have KDE Plasma 5 and the Android device has Cyanogenmod 12.1 by the way. If I look into the battery usage overview kdeconnect is not listed so that it probably is under the one percent threshold – it even occurs to have no partial wakelocks as I can examine with BetterBatteryStats. Is there anything I can try to aid in this problem?
*** Bug 339406 has been marked as a duplicate of this bug. ***
Doesn't seem to happen anymore.