I'm curious how many people run Ziti full time on a mobile device (Android and IOS) and if it seriously affects battery runtime. Looks like the last post on battery life was 2022. I did some testing and found it was a noticeable drain on my Pixel 6a. Obviously mobile device OS + transparent network stack + individual overlay configuration can be an endless number of interacting complex variables. But still I wonder what's the current state of knowledge on this and/or if there's any "pro tips" to minimize draining of users' devices.
It is on our radar still.
The SDKs keep themselves awake due to a polling process that we are eliminating with a new feature set that allows push events. Once released and in use, we believe this will improve battery life.
If we still encounter battery issues after that is removed, we will have to dive deeper into how the SDKs are working with the OS to avoid as many wakes as possible.
In addition to the polling vs eventing update @andrew.martinez mentions, we are also also updating the Android tunnelers to use the same underpinning as the iOS tunneler, which we expect improve its battery usage and overall stability. (Android seems to have battery issues at a much higher rate than Ziti Mobile Edge for iOS. A smaller subset of users see battery issues on iOS)
We have made attempts to go into "quiet mode" based on device events, but abandoned these efforts due to poor results. This issue was created to track an alternate approach we have discussed.