Segfault > IT > Hardware > Smartphone Intex Aqua Fish running Sailfish OS > Sailfish OS - Battery life (on the Intex Aqua Fish)
First published on: 19.Oct.2016

Last change: 19.Dec.2016

Battery life

Sailfish OS on Intex Aqua Fish

With the "Intex Aqua Fish" the situation is kind of grim.

No matter what I analyzed or did, the baseline seems to be that unless the device is switched off, the battery depletion when doing "nothing" (no apps running, all Intex-stuff uninstalled, no SIM-card, device in flight-mode, no antennas active, OS version is 1% per hour, which is A LOT.

The 1% per hour means basically that the phone will hold AT MOST 4 days even if you don't use nor touch it and have it disconnected from the network.

I'm not sure what to think:

  • I do see many debugging options switched on in the kernel ("less /proc/config.gz"), but on the other hand the sleep-state-analysis reports that the CPU is basically sleeping all the time, so this in turn might mean that those debug options do not put the OS under pressure or on the other hand it might be that the CPU could be able to sleep even more once the debug-options are switched off?
  • any phone that runs Sailfish OS has to keep all the time an eye on the display surface (because of the double-tap functionality to wake it up) => how much battery does that need?

I might try to compile during the next weeks a kernel that has all debugging feats turned off to then see if it changes anything... .

Here are some stats while leaving the phone idle on a table, with all antennas and receivers switched off:

battery charge

cpu sleep

cpu idle

network traffic

cpu used frequencies

Sailfish OS itself

While using the phone the battery drains less fast than on my Huawei P9 - great!

I did fiddle around with other CPU-schedulers ("ondemand", "interactive"+other settings) and I ended up thinking that the current pre-set "interactive" is a good setting, but anyway with the 1% per hour battery consumption of my current device as baseline (Intex Aqua Fish) any additional optimizations fade into noise.