nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Child Item

Datastore sync effects on temperature/energy

Caram

During testing this morning on an iPhone, I notice the device's temperature was keeping unusually cool. The iPhone was in silent mode. When I performed this kind of testing in the past, network was always on and the device's temperature rapidly became noticeably warm/hot and the battery was draining rapidly. No such occurrence in silent mode.

Is there a known issue with uploading deltas in terms of energy efficiency? Are there any recommendations to reduce the energy drain/temperature increase? Obviously turning sync off is not an option. ;-)

Caram

Find more posts tagged with

API

Abuse

Comments

Steve M.

Were you doing something active with datastores?

In terms of network usage: as long as you have an observer, the Datastore SDK will keep a single long-polling HTTP connection open with the server and wait for changes. Every few minutes this connection is closed and reestablished. If you're actively making changes and syncing them, HTTP connections are being made to the server (perhaps as often as once per delta), and those deltas are transmitted.

In terms of battery life, using the network can certainly be a contributor, but it's also possible there's local processing that's burning CPU cycles. Again, if no active use is happening, this should be really minimal, but if you're making a lot of changes and syncing them, there will be some CPU usage.

I'm not clear on what "silent mode" has to do with this... I wouldn't expect that to interact with the datastore use at all?

Caram

Yes, plenty of active work, creating and deleting datastores, create and deleting tables, creating and deleting records, but not too many of them each time but fairly regularly (and yes, I have at least 2 observers active at once). The only difference with my normal testing sessions is that "silent mode" was on, so no network traffic could occur. This leads me to suspect that the heat I usually feel is due to heavy network traffic (and possibly the CPU processing by the background Datastore SDK threads). If this was purely due to heavy processing in the app, the CPU would heat up also in "silent mode".

Have you ever done any energy profiling on the SDK using Xcode 6 ? It might be worth a try.

Steve M.

Sure, I would imagine that much of the energy use is due to network traffic.