Using App Metrics

You can think of workload as the sum of all the work performed by a variety of activity types.

The App Metrics dashboard gives you several visualizations of the work that your app is doing. The charts serve two purposes:

• To give you an overview of your app’s total workload

• To give you an easy way to drill down into each metric category and get a granular view of how single workflows and expressions contribute

Metric categories

Each of the metric categories and how they can be explored in the App Metrics dashboard is listed in the table below. Click each category type to learn more about it.

Workload usage

The bar graph helps you get a transparent view of how much app your work is doing in total, and at what times. This visualization lets you see trends and patterns in your app's workload usage over time, which can help you plan and make more informed decisions about future upgrades or changes to your app.

Each column in the bar graph represents one day, going back 30 days from the current date by default. This example above is from a complex app that processes thousands of users, searches and workflows.

Changing the time frame

To drill down into an individual day, simply click that day in the chart. Let’s look at August 22nd when the workload was a bit higher:

Here, we can see that something occurred between 6:00 pm (18:00) and 9:00 pm (21:00) that increased the workload for the day.

Clicking it allows you to zoom in on that specific time period to better understand what caused the spike in activity and added extra tasks for Bubble to handle.

Isolating the timeframe extends down to the granular pie chart below, which lets you identify activity types on the top level and drill down into individual workflows and expressions to see what they are doing.

Granular view

The granular view gives you a pie chart where each slice represents one of the activity types that we track. This way, you get an understanding of the types of processes that contribute the most.

Let’s look at an example:

In the chart above we can see clearly that two activities make up the majority of the work that this app performs:

• Scheduled workflows

• Data exports

Hovering that section in the pie chart or the legend on the right side will provide more information about the actual number of WU spent, and its percentage of the total.

This data alone can already tell us something about how the app is built: Scheduled workflows are server-side API workflows, and they make up a singificant part of the total in this example. A large part also goes to the Data export feature, suggesting that one or more large exports have taken place in this time period.

Activity details

By hovering one of the slices in the pie chart (or one of its corresponding entries in the legend on the right side), you get the following additional information:

• Total WU consumption

• Percentage of total consumption

• Number of activity runs

Drilling down

Clicking on each of these activity types provides an even more detailed view of individual instances of that activity type, and how much each one contributes to its total. Let’s click on the Scheduled workflow sector to dig deeper:

Now we are down to a list of different data fetching operations.

By again hovering the mouse over one of the vectors in the chart, we can see the percentage consumed by that specific activity. Clicking that takes us directly to the workflow in the workflow editor that's consuming the workload, so that we can investigate it.

This way of continually drilling down in the chart can teach us a lot about where the work happens and where we can potentially make adjustments to make the app more efficient.

This not only helps you keep your app operating at a cost-effective level, but it can also speed up processes to improve the overall experience for users.

Excluding details

As you research your app’s workload, it can sometimes be useful to see what the chart looks like if we exclude some information. For example, if you want to stop focusing on Scheduled workflows, you can hide that information in the chart by clicking on that activity type in the right-hand list:

As you can see, when we click on the activity entry in the list, it becomes crossed out, and the chart is updated to exclude it: the dark blue slice is no longer visible.

Real-time metrics

The real-time workload reporting allows you to view and access workload data in real time. The bar chart provides one-minute granularity for the last 24 hours. This is useful for checking recent or highly time-specific workloads to learn more about how they contribute in isolation.

Checking specific workload units consumed in the logs

Using the Server logs tab, you can check the workload charged for each action, as well as the total for each workflow.

Using the data

Creating efficient applications is a process of continuous improvement. Using the App Metrics dashboard is an effective method for recognizing the processes that have the greatest impact on your app’s workload. As you analyze your workload data, you will discover that some activities are essential for your app to operate, while others can be optimized.

The flexibility of Bubble’s platform opens up to different ways of solving the same problem, and sometimes small changes can lead to great improvements in efficiency.