But much of the process of application startup happens, by definition, before the system gets around to calling your cod

Performance testing should always involve running a test many times, to reduce the inherent variability in results; the more runs you do, the more faith you can put in the averaged results. I try to run tests a minimum of ten times, but doing even more is even better. Depending on how variable the results are in your situation, and how small the timings are (since variability would have larger impact on smaller durations), doing many more runs might be necessary.

This duration (1,380ms in this example) represents the time that it took from launching the app to the time when the system consider it “launched,” which includes drawing the first frame (hence “Displayed”).

This Displayed duration does not necessarily include everything your app needs to do before it’s ready to go. You can supply that extra information to the system by calling Activity.reportFullyDrawn() whenever your app determines it’s completely done loading and initializing. If/when you call that optional method, the system issues another log with that timestamp and duration:

To convert the output above to just the launch duration, I pipe the output through grep and cut shell commands (there are various ways to do this, I just picked one at random):

The last argument is the package component information for your app. You can see that they are the same here as in the ActivityTaskManager log output in the section above.

I poked around in logcat, looked at some low-level APIs, and asked some of the engineers on the platform team, and found a few things to help. Better yet, I was able to use facilities in the adb shell tool to completely automate my testing and output information in a way that made it easy to pop the results into a spreadsheet and analyze them.

So what I really wanted was some way to launch the app from the command line. Then I could run that command to do the same thing over and over, and avoid the variability (and tediousness) of manually launching the app by hand.

This means that we don’t have to troll through logcat to get this information. Instead, we can get it directly from the console where we run the launch command. Better yet, we can strip the extraneous text and leave just the launch result, making it easier to extract this data for use elsewhere.

Running this command launches the app (unless the app is already in the foreground, which is not what you want; we’ll deal with that next), and then outputs the following information:

Tapping on an app icon to launch it many times in a row is… pretty tedious. And doing so in a predictable and consistent way (to avoid introducing variability, like if you happened to launch another app by mistake, or otherwise made the system do extra work throwing off timing results) can be a problem.

Both scenarios are valid to test and understand. But in general, testing your startup performance in a cold start scenario is a better place to start, for two main reasons:

Starting: Intent { cmp=com.android.samples.mytest/.MainActivity } Status: ok LaunchState: COLD Activity: com.android.samples.mytest/.MainActivity TotalTime: 1380 WaitTime: 1381 Complete

adb (Android Debug Bridge, a tool probably familiar to anyone who’s read this far) provides just the thing I needed. More specifically, adb shell provides a command-line interface to launch an app: adb shell am start-activity. The command should also block until the app has finished launching, so we’ll use the -W argument as well (this is necessary for the next step, where we’ll tack on a follow-up command to kill the app post-launch). Here’s the complete launch command:

As I wrote before in an earlier (and unfortunately outdated and incorrect) blog, there’s a handy log that the system has been issuing since the KitKat release. Whenever an activity starts, you’ll see something like this in the logcat output:

http://old.cocir.org/media/qas/Video-odense-bulldogs-v-frederikshavn-white-hawks-v-da-da-1gpf-2.php

http://molos.bodasturias.com/qxz/Video-tecnyconta-zaragoza-v-nizhny-novgorod-bc-v-es-es-1ntk-8.php

http://www.ectp.org/kzz/videos-v-gr-gr-1myc-9.php

http://team.vidrio.org/xpy/Video-KS-Cracovia-Krakow-Unia-Oswiecim-v-en-gb-1ukv-3.php

http://startup.munich.es/mlt/videos-gmac-bologna-v-pinar-karsiyaka-v-tr-tr-1hjm-2.php

http://old.cocir.org/media/qas/video-odense-bulldogs-v-frederikshavn-white-hawks-v-da-da-1hrz-18.php

http://main.ruicasa.com/tgq/videos-Bili-Tygri-Liberec-HC-Vitkovice-Ridera-v-en-gb-ywg30122020-.php

http://team.vidrio.org/xpy/v-ideos-Kometa-Brno-Aukro-Berani-Zlin-v-en-gb-1qdp-.php

http://old.cocir.org/media/qas/video-Lulea-Hockey-Rogle-BK-v-en-gb-1qfl30122020-.php

http://team.vidrio.org/xpy/Video-Kometa-Brno-Aukro-Berani-Zlin-v-en-gb-1thb-11.php

http://elta.actiup.com/cqn/videos-Norway-Belarus-v-en-gb-glh-.php

http://startup.munich.es/mlt/videos-gmac-bologna-v-pinar-karsiyaka-v-tr-tr-1hbh-26.php

http://main.ruicasa.com/tgq/videos-Bili-Tygri-Liberec-HC-Vitkovice-Ridera-v-en-gb-wvw-.php

http://molos.bodasturias.com/qxz/v-ideos-tecnyconta-zaragoza-v-nizhny-novgorod-bc-v-es-es-1noh-6.php

http://old.cocir.org/media/qas/v-ideos-Lulea-Hockey-Rogle-BK-v-en-gb-1pqp-5.php

http://elta.actiup.com/cqn/video-Norway-Belarus-v-en-gb-upj30122020-.php

http://main.ruicasa.com/tgq/video-Bili-Tygri-Liberec-HC-Vitkovice-Ridera-v-en-gb-fnv30122020-.php

http://elta.actiup.com/cqn/Video-Norway-Belarus-v-en-gb-qfy-.php

http://startup.munich.es/mlt/video-gmac-bologna-v-pinar-karsiyaka-v-tr-tr-1hfp-12.php

http://old.cocir.org/media/qas/v-ideos-Lulea-Hockey-Rogle-BK-v-en-gb-1ujp-19.php

o drop columns with missing values. We can also set a threshold value for the number of non-missing values required for a column or row to have. For instance, thresh=5 means that a row must have at least 5 non-missing values not to be dropped. The rows that have 4 or fewer missing values will be dropped.