As mentioned in a previous thread [1], we still use the 319MB Flame
configuration as it's the only comparable baseline we have between v2.2 and
v2.5. We can't drop using the 319MB Flame for v2.5 unless we build a new
baseline with a different device configuration against v2.2.

[1]
https://groups.google.com/d/msg/mozilla.dev.fxos/uDrAUHGTq18/79YaFJHQCQAJ

Thanks,

Eli Perelman

On Fri, Oct 2, 2015 at 12:58 PM, Justin D'Arcangelo <[email protected]
> wrote:

> Correction, my bad. Eli just informed me that his numbers earlier were
> from the 319mb Flame configuration. I was under the impression that we were
> no longer supporting that config. In addition, I was looking at
> “fullyLoaded” and not “visuallyLoaded”.
>
> We have a patch on the way that should be optimizing the way we fetch
> album art and should drastically cut down on memory usage. Once that has
> landed, I will re-test under 319mb and see where we stand.
>
> -Justin
>
>
> On Oct 2, 2015, at 1:51 PM, Justin D'Arcangelo <[email protected]>
> wrote:
>
> Here’s the latest Raptor numbers, run locally, on Flame:
>
> OGA:
>
> jdarcangelo-20869:gaia Justin$ raptor test coldlaunch --app music-oga
> --runs 5
> [Cold Launch: music-oga.gaiamobile.org] Preparing to start testing...
> [Cold Launch: music-oga.gaiamobile.org] Priming application
> [Cold Launch: music-oga.gaiamobile.org] Starting run 1
> [Cold Launch: music-oga.gaiamobile.org] Run 1 complete
> [Cold Launch: music-oga.gaiamobile.org] Starting run 2
> [Cold Launch: music-oga.gaiamobile.org] Run 2 complete
> [Cold Launch: music-oga.gaiamobile.org] Starting run 3
> [Cold Launch: music-oga.gaiamobile.org] Run 3 complete
> [Cold Launch: music-oga.gaiamobile.org] Starting run 4
> [Cold Launch: music-oga.gaiamobile.org] Run 4 complete
> [Cold Launch: music-oga.gaiamobile.org] Starting run 5
> [Cold Launch: music-oga.gaiamobile.org] Run 5 complete
> [Cold Launch: music-oga.gaiamobile.org] Results from
> music-oga.gaiamobile.org
>
> | Metric                | Mean     | Median | Min    | Max    | StdDev |
> p95      |
> | --------------------- | -------- | ------ | ------ | ------ | ------ |
> -------- |
> | navigationLoaded      | 737      | 739    | 715    | 752    | 14.071 |
> 737      |
> | navigationInteractive | 900.600  | 893    | 880    | 924    | 16.305 |
> 900.600  |
> | visuallyLoaded        | 1123     | 1121   | 1076   | 1170   | 32.323 |
> 1123     |
> | contentInteractive    | 1123.400 | 1121   | 1077   | 1171   | 32.327 |
> 1123.400 |
> | fullyLoaded           | 2478.400 | 2474   | 2450   | 2517   | 23.466 |
> 2478.400 |
> | uss                   | 14.542   | 14.645 | 14.301 | 14.738 | 0.182  |
> 14.542   |
> | rss                   | 34.377   | 34.480 | 34.133 | 34.574 | 0.185  |
> 34.377   |
> | pss                   | 18.775   | 18.877 | 18.534 | 18.978 | 0.186  |
> 18.775   |
>
> [Cold Launch: music-oga.gaiamobile.org] Testing complete
> jdarcangelo-20869:gaia Justin$
>
>
> NGA:
>
> jdarcangelo-20869:gaia Justin$ raptor test coldlaunch --app music --runs 5
> [Cold Launch: music.gaiamobile.org] Preparing to start testing...
> [Cold Launch: music.gaiamobile.org] Priming application
> [Cold Launch: music.gaiamobile.org] Starting run 1
> [Cold Launch: music.gaiamobile.org] Run 1 complete
> [Cold Launch: music.gaiamobile.org] Starting run 2
> [Cold Launch: music.gaiamobile.org] Run 2 complete
> [Cold Launch: music.gaiamobile.org] Starting run 3
> [Cold Launch: music.gaiamobile.org] Run 3 complete
> [Cold Launch: music.gaiamobile.org] Starting run 4
> [Cold Launch: music.gaiamobile.org] Run 4 complete
> [Cold Launch: music.gaiamobile.org] Starting run 5
> [Cold Launch: music.gaiamobile.org] Run 5 complete
> [Cold Launch: music.gaiamobile.org] Results from music.gaiamobile.org
>
> | Metric                | Mean     | Median | Min    | Max    | StdDev |
> p95      |
> | --------------------- | -------- | ------ | ------ | ------ | ------ |
> -------- |
> | navigationLoaded      | 804.200  | 808    | 742    | 858    | 37.151 |
> 804.200  |
> | navigationInteractive | 826.400  | 827    | 773    | 879    | 33.643 |
> 826.400  |
> | visuallyLoaded        | 1655.600 | 1625   | 1597   | 1788   | 68.002 |
> 1655.600 |
> | contentInteractive    | 1656     | 1625   | 1598   | 1789   | 68.220 |
> 1656     |
> | fullyLoaded           | 1656.600 | 1625   | 1598   | 1790   | 68.482 |
> 1656.600 |
> | uss                   | 20.751   | 20.855 | 19.855 | 21.309 | 0.491  |
> 20.751   |
> | pss                   | 24.955   | 25.103 | 24.015 | 25.555 | 0.518  |
> 24.955   |
> | rss                   | 40.479   | 40.594 | 39.578 | 41.035 | 0.493  |
> 40.479   |
>
> [Cold Launch: music.gaiamobile.org] Testing complete
> jdarcangelo-20869:gaia Justin$
>
>
> I’m actually seeing ~800ms *improvement* with the NGA app in
> “fullyLoaded”. Maybe Eli can answer this, but is “fullyLoaded” the number I
> should be looking at? I’m also only seeing a ~6MB regression in USS.
> All-in-all, this doesn’t seem “that bad” for what is essentially an
> unoptimized app. Again, I’m confident that we can improve upon these
> numbers significantly in the next 2-3 weeks.
>
> It also looks like the Gecko-wide perf regression is now gone (I hope).
> So, maybe we can start trusting these numbers again.
>
> -Justin
>
>
> On Oct 2, 2015, at 1:16 PM, Fabrice Desré <[email protected]> wrote:
>
> On 10/02/2015 09:49 AM, Justin D'Arcangelo wrote:
>
> I would also like to add that this policy of immediately pouncing on devs
> who attempt to try something new that may cause the perf numbers to
> momentarily dip is part of why we seem to have a culture problem in FxOS
> dev where everyone is afraid to take any kind of risks. If we are not
> allowed to have a 2-3 week window to optimize after a huge landing such as
> this, then how are we supposed to experiment or take risks?
>
>
> You have all the time you want if you don't put dogfooders at risk. No
> one is saying that you should not take the risk to try something new
> (side note, you spent enough time on spark & flyweb to know that). But
> when it comes to shipping there is a minimum bar to meet, and with
> basically a x2 memory usage we are not meeting it in this app yet,
> sorry. Feel free to ship a new app alongside the existing one instead
> and ask people to try it, since we can't do A/B testing.
>
> The problem we have is that most people don't care enough about having a
> stable nightly, which is why we haven't updated dogfooders for more than
> a month now.
>
> Fabrice
> --
> Fabrice Desré
> b2g team
> Mozilla Corporation
>
>
>
>
> _______________________________________________
> dev-fxos mailing list
> [email protected]
> https://lists.mozilla.org/listinfo/dev-fxos
>
>
_______________________________________________
dev-fxos mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-fxos

Reply via email to