Measuring the Execution Time of Liquid Code

Prerequisites

This guide is an advanced tutorial. To follow it, you must understand basic Insites concepts such as the topics in the Get Started section, the and Liquid filters and how to create application logs.

Steps

Measuring the execution time of Liquid code fragments using the Liquid filter is a three-step process:

• Step 1: Identifying measuring points
• Step 2: Adding the 'to_time' and 'time_diff' filters
• Step 3: Checking results in the logs

Step 1: Identifying measuring points

Locate the part of the code you want to measure. Take a look at this sample snippet:

{% graphql g = "modules/n_plus_one/n1_programmers" %}

{% for programmer in g.programmers.results %}
  <tr>
    <td>{{ programmer.properties.email }}</td>
    <td>{{ programmer.properties.title }}</td>
    <td>{{ programmer.company.properties.name }}</td>
  </tr>
{% endfor %}

Two items could potentially cause results to be displayed slowly:

1. GraphQL query
2. for loop rendering records

Step 2: Adding the 'to_time' and 'time_diff' filters

Use the and filters to measure execution times.
The and variables store partial execution times in milliseconds. The variable stores the overall execution time of the code block.

{% liquid
  assign __start = "now" | to_time
  graphql g = "modules/n_plus_one/n1_programmers"
  assign __graph = "now" | to_time
%}

{% for programmer in g.programmers.results %}
  <tr>
    <td>{{ programmer.properties.email }}</td>
    <td>{{ programmer.properties.title }}</td>
    <td>{{ programmer.company.properties.name }}</td>
  </tr>
{% endfor %}

{% liquid
  assign __render = "now" | to_time
  assign _total_duration = __start | time_diff: __render
  assign _graph_duration = __start | time_diff: __graph
  assign _render_duration = __graph | time_diff: __render
  log _total_duration, type: "benchmark-total"
  log _graph_duration, type: "benchmark-graph"
  log _render_duration, type: "benchmark-render"
%}

Step 3: Checking results in the logs

Test results are sent to your instance log stream and can be accessed by using the tool.

Next Steps

Congratulations! You now know how to measure the execution time of Liquid code fragments using the filter. You may visit another tutorial that uses this method for testing two different solutions: Loading Related Records