Trace is a Bitrise add-on designed to monitor your applications from the perspective of its end users. It comes with out-of-the-box performance monitoring, giving you insight into how the app is performing across the entire install base.
The core philosophy of Trace is to simplify how mobile product teams:
- Detect the problem: know about issues before your users report them.
- Assess the impact: focus on resolving the issues that have the greatest impact on your users.
- Trace the cause: spend less time trying to reproduce issues.
Accessing Trace on Bitrise ⚓
To access the Trace add-on on Bitrise:
- Open an app on Bitrise.
- Go to the Add-ons tab.
- Click Go to add-on.
However, to actually use Trace and being able to gather data from it, you’ll need the Trace SDK installed to your app.
Installing Trace ⚓
Trace works by installing the Trace SDK to your application during the Bitrise build process. You can do this manually on any device to test it and once you’re ready to automate your process, you can simply use our dedicated Step on Bitrise.
Installing Trace manually ⚓
Before adding Trace to your apps on Bitrise, try it out first. We’ll walk you through the process using a sample app - but of course you can use any of your own repositories instead of our sample!
In this example, we’re going to use CocoaPods to install the Trace SDK. Let’s see how!
Please note that once you installed Trace, it can take up to an hour for it to start process your data - until then you won’t see any data on your Trace dashboard.
-
Clone the app and go to its directory on your machine:
$ git clone https://github.com/bitrise-io/BookStore-iOS.git $ cd BookStore-iOS/ -
Open the Podfile of the app and add the Trace SDK to it:
# Podfile target 'YOUR_TARGET_NAME' do pod 'BitriseTrace' end - Install the SDK by running pod install.
If you’ve done everything right, the Pod installation complete! message should appear. - Open Trace on Bitrise.
- In the top-right, click Getting started.
- Find the Download Config file step, and click Download.
-
Open your app in Xcode:
$ open BookStore.xcworkspace/ -
Add the bitrise_configuration.plist file to your app by dragging and dropping it to the root of the project.
-
If prompted to choose from different options for adding the files, choose your project - BookStore in our example -, and click Finish.
- Run your app in a simulator or on a device!
Installing Trace on Bitrise ⚓
On Bitrise itself, it’s much simpler to install Trace: you simply need to add the Add Trace SDK Step to your Workflow. The Step must come before the application’s binary itself is built. For example, if you build an iOS app, the Step must come before the Xcode Archive & Export for iOS Step.
To install Trace:
- Open the Workflow Editor of your app.
- Open the Workflow you use to build the app.
- Add the Add Trace SDK Step before the Step that builds your’s app binary.
- Check that the Project (or Workspace) path and the Scheme name input values are both correct.
By default, both values are Environment Variables which are created and stored at the creation of the app. - Run a build!
If all went well, Trace will track application data. Let’s talk about how to analyze that data that Trace provides!
Analyzing application performance ⚓
The first stop for analyzing performance is the Performance page of Trace. The primary purpose of the Performance page is to give you an overview of the performance of your app across the app’s entire install base.
You can check the performance of your apps in three distinct categories of metrics on the Performance page:
- Application stability metrics: they show how reliable your app is.
- Application responsiveness metrics: they show how quickly the app responds to user interaction.
- Application consumption metrics: they show how much resources your app needs to run.
On the Performance page, you can check all these metrics, and break them down based on various properties, such as device type, operating system, network carrier, and so on. Read on to find out about the specific metrics, and how to take advantage of Trace’s analysis features.
Application stability metrics ⚓
Trace tracks three stability-related metrics:
- Crashes: how many fatal errors the app has reported.
- Errors: how many non-fatal errors have been generated.
- Network errors: how many network request errors have been generated.
Application responsiveness metrics ⚓
Trace tracks three responsiveness-related metrics:
- Cold start latency: how long the app takes to load when it is already running in the background.
- Warm start latency: how long the app takes to be ready when returning to the foreground from a background state.
- Network request latency: how long network requests made by the app take to complete.
Application consumption metrics ⚓
Trace tracks three metrics related to what an application consumes:
- CPU: this the total CPU usage of the the app across all cores.
- Memory: the total number of bytes of memory actively used by the app.
- Request count: this is the number of network requests being made by the app.
Filtering and sorting metrics ⚓
On the Trace Performance page, you can:
- Filter the tracked sessions to focus only on certain session types in your analysis.
- Get a breakdown of the metrics with sorting them based on certain properties.
- Set a data range for the metrics you want to view.
Various session properties are available for filtering, including:
- App version: The version of the application.
- Country: The country the session originated from.
- OS version: The version of the device’s operating system.
- Device type: The type of device on which the session ran.
- Network: The type of network the app is connected from.
- Carrier: The cellular carrier for the device.
You can filter for any and all of these. For example, you can set the Country filter to United States and the Device filter to iPhone X. This means that Trace will show the metrics collected from users in the United States who use your app on an iPhone X.
By default, no filter is active: the metrics of all sessions monitored by Trace are displayed on the Performance page. To set a filter:
- Click on the filter bar below the Performance heading.
By default, this bar displays All sessions. - Choose the filters you need.
You can also group the individual metrics based on the above categories. Grouping means getting a breakdown of the metric based on the selected category. To do so:
- Find the metric you need; for example, App Error Rate.
- To the right of the name of the metric, open the dropdown menu.
- Choose the category you need.
For example, if you select Device, you will see the app error rate, grouped by device type.
You can also set a date range for the available metrics. To do so:
- Click the date range dropdown menu in the top right (by default, it says Last 24 Hours).
- Choose the date range you need.
Checking the details of a metric ⚓
To check any of the metrics:
- Go to the Performance page.
- Find the metric you need.
- Click Details to go to the Performance details page.
Analyzing performance details ⚓
The Performance details page gives you a full breakdown of any selected metric. You can apply the same filtering and sorting to the sessions as on the Performance page, and set a date range the same way.
At the top of the page, you can see the metric inspector: it is a chart that displays the metric’s data over the specified time period, which you can set in the dropdown menu found in the top right corner. Apply filters and sorting to narrow down your analysis of any issues.
Below the metric inspector, you’ll see detailed breakdowns of all session properties, such as app version, country, network carrier, and so on. For each breakdown you can see a plot for each item of that property. To put it in simpler terms: in the App versions breakdown of, say, the Network errors metric, you will see plots for the network errors of each app version that is tracked by Trace. You can easily create a plot analysis:
- Find the breakdown you need (for example, App versions).
- Click the specific plot you need (for example, version 2.6.30).
This will apply a session filter. It is the same as setting the same filter in the Session filter bar at the top of the page.
And that’s it: now every breakdown will show data related to only the selected plot (in our example, app version 2.3).
Analyzing sessions ⚓
Performance metrics in themselves give no indication to what the users of the apps did in a specific session. For example, the app crash rate shows the number of crashes but doesn’t give you any context to the crashes themselves, and how the users triggered them. To do so, you need to use the Session Explorer page.
The Session Explorer shows the specific details of what happened during a user session as well as details about the environment on which the session occurred, all vital in locating and reproducing a problem. The data includes all important session properties, such as app version, device type, OS version, network type, and carrier, as well as the duration of the session and a user journey. The user journey shows what parts of the app the user viewed and interacted with and the duration of the interactions.
To check out individual sessions:
- Go to the Performance page in Trace.
- Select a metric - for example, App Crash Rate -, and click Details.
- In the top right corner, click See User Sessions.
- Set a date range for the sessions you want to see - for example, 24 hours.
- Navigate between the different user sessions with the in the top right.
- To see a user journey in detail, scroll down to the User Journey section, and click Show All.