Performance and Scalability for Drupal 8, 9, and 10

Drupal comes with all of its caching features enabled by default. This improves response time, but can be frustrating for themers as it makes it harder to preview the changes you make to template files.

In this tutorial we'll look at:

• Why these features are enabled by default
• How the theme layer leverages Drupal's caches
• Why you should learn to disable them when doing development

The Render API is capable of detecting poorly-cacheable (highly dynamic) parts of a page and rendering them later using a process called auto-placeholdering. This works by using #lazy_builder callbacks to lazy load certain very dynamic subtrees of a render array. The place in the array where that very dynamic content would appear is first assigned a placeholder. At the very last moment it is replaced with the actual content.

This allows Drupal to do things like cache the overall page in the Dynamic Page Cache despite parts of the page being too dynamic to be worth caching. It also allows the Render API to assemble a page using cache fragments combined with non-cacheable elements.

In this tutorial we'll:

• Discuss what lazy builders are and how they work in conjunction with placeholders to speed up the rendering pipeline
• Cover some common gotchas for lazy builders
• Look at some example code that implements a lazy builder callback

By the end of this tutorial, you should know how and when to use the #lazy_builder property of a render array and how Drupal uses placeholders to increase the cacheability of content and speed up the rendering process.

The individual items that make up the content of a page impact the cacheability of that page. In order for Drupal's cache and external caches to better understand how the content varies on a page, module developers use the #cache render element property. The #cache property defines cacheability metadata for individual elements in a render array.

Additionally, these Render API elements can become fairly complex. The calculation of what the final HTML output should look like often involves looking up content in the database, checking multiple conditions, maybe querying an external API, and various other tasks. This can cause turning a render array into HTML to become quite expensive. In order to speed up this process, the Render API will cache the generated HTML for each element and reuse it on future requests whenever possible -- but only if you tell it to do so.

In this tutorial, we'll look at:

• How render caching impacts the performance of a page
• Defining the cacheability of an item with cache tags, cache contexts, and cache max-age
• Examples of using the #cache property in a render array

By the end of this tutorial you should know how, and when, to use the #cache property when defining render arrays.