/longtasks

Long Task API

Primary LanguageBikeshedOtherNOASSERTION

Long Task API

Long Tasks is a new real user measurement (RUM) performance API to enable applications to measure responsiveness. It enables detecting presence of “long tasks” that monopolize the UI thread for extended periods of time and block other critical tasks from being executed - e.g. reacting to user input.

Background

As the page is loading and while the user is interacting with the page afterwards, both the application and browser queue various events that are then executed by the browser -- e.g. the user agent schedules input events based on user’s activity, the application schedules callbacks for requestAnimationFrame and other callbacks etc. Once in the queue, these events are then dequeued one-by-one by the browser and executed — see “the anatomy of a frame” for a high-level overview of this process in Blink.

However, some tasks can take a long time (multiple frames), and if and when that happens, the UI thread is locked and all other tasks are blocked as well. To the user this is commonly visible as a “locked up” page where the browser is unable to respond to user input; this is a major source of bad user experience on the web today:

  • Delayed “time to Interactive”: while the page is loading long tasks often tie up the main thread and prevent the user from interacting with the page even though the page is visually rendered. Poorly designed third-party content is a frequent culprit.
  • High/variable input latency: critical user interaction events (tap, click, scroll, wheel, etc) are queued behind long tasks, which yields janky and unpredictable user experience.
  • High/variable event handling latency: similar to input, but for processing event callbacks (e.g. onload events, and so on), which delay application updates.
  • Janky animations and scrolling: some animation and scrolling interactions require coordination between compositor and main threads; if the main thread is blocked due to a long task, it can affect responsiveness of animations and scrolling.

Some applications (and RUM vendors) are already attempting to identify and track cases where “long tasks” happen. For example, one known pattern is to install a ~short periodic timer and inspect the elapsed time between the successive calls: if the elapsed time is greater than the timer period, then there is high likelihood that one or more long tasks have delayed execution of the timer. This mostly works, but it has several bad performance implications: the application is polling to detect long tasks, which prevents quiescence and long idle blocks (see requestIdleCallback); it’s bad for battery life; and there is no way to know what caused the delay. (e.g. first party vs third party code)

RAIL performance model suggests that applications should respond in under 100ms to user input; for touch move and scrolling in under 16ms. Our goal with this API is to surface notifications about tasks that may prevent the application from hitting these targets.

Terminology

Major terms:

  • frame or frame context refers to the browsing context, such as iframe (not animation frame), embed or object
  • culprit frame refers to the frame or container (iframe, object, embed etc) that is being implicated for the long task
  • attribution refers to identifying the type of work (such as script, layout etc.) that contributed significantly to the long task AND which browsing context or frame is responsible for that work.
  • minimal frame attribution refers to the browsing context or frame that is being implicated overall for the long task

V1 API

Long Task API introduces a new PerformanceEntry object, which will report instances of long tasks:

interface PerformanceLongTaskTiming : PerformanceEntry {
  [SameObject, SaveSameObject] readonly attribute FrozenArray<TaskAttributionTiming> attribution;
};

Attribute definitions of PerformanceLongTaskTiming:

  • entryType: "longtask"

  • startTime: DOMHighResTimeStamp of when long task started

  • duration: elapsed time (as DOMHighResTimeStamp) between start and finish of task

  • name: minimal frame attribution, eg. "same-origin", "cross-origin", "unknown" etc. Possible values are:

  • "self"

  • "same-origin-ancestor"

  • "same-origin-descendant"

  • "same-origin"

  • "cross-origin-ancestor"

  • "cross-origin-descendant"

  • "cross-origin-unreachable"

  • "multiple-contexts"

  • "unknown"

  • attribution: sequence of TaskAttributionTiming, a new PerformanceEntry object to report attribution within long tasks. To see how attribute is populated for different values of name see the section below: Pointing to the culprit

interface TaskAttributionTiming : PerformanceEntry {
  readonly attribute DOMString containerType;
  readonly attribute DOMString containerSrc;
  readonly attribute DOMString containerId;
  readonly attribute DOMString containerName;
};

Attribute definitions of TaskAttributionTiming:

  • entryType: “taskattribution”
  • startTime: 0
  • duration: 0
  • name: type of attribution, eg. "script" or "layout"
  • containerType: type of container for culprit frame eg. "iframe" (most common), "embed", "object".
  • containerName: DOMString, container’s name attribute
  • containerId: DOMString, container’s id attribute
  • containerSrc: DOMString, container’s src attribute

Long tasks events will be delivered to all observers (in frames within the page or tab) regardless of which frame was responsible for the long task. The goal is to allow all pages on the web to know if and who (first party content or third party content) is causing disruptions.

The name field provides minimal frame attribution so that the observing frame can respond to the issue in the proper way. In addition, the attribution field provides further insight into the type of work (script, layout etc) that caused the long task as well as which frame is responsible for that work. For more details on how the attribution is set, see the "Pointing to the culprit" section.

The above covers existing use cases found in the wild, enables document-level attribution, and eliminates the negative performance implications mentioned earlier. To receive these notifications, the application can subscribe to them via PerformanceObserver interface:

const observer = new PerformanceObserver(function(list) {
  for (const entry of list.getEntries()) {
     // Process long task notifications:
     // report back for analytics and monitoring
     // ...
  }
});


// Register observer for long task notifications.
// Since the "buffered" flag is set, longtasks that already occurred are received.
observer.observe({type: "longtask", buffered: true});

// Long script execution after this will result in queueing 
// and receiving “longtask” entries in the observer.

Long-task threshold is 50ms. That is, the UA should emit long-task events whenever it detects tasks whose execution time exceeds >50ms.

Demo

For a quick demo, visit this website on a browser which supports the Long Tasks API.

For a demo of long tasks from same-origin & cross-origin frames, see this website. Interacting with the iframed wikipedia page will generate cross-origin long task notifications.

Pointing to the culprit

Long task represents the top level event loop task. Within this task, different types of work (such as script, layout, style etc) may be done, and they could be executed within different frame contexts. The type of work could also be global in nature such as a long GC that is process or frame-tree wide.

Thus pointing to the culprit has couple of facets:

  • Pointing to the overall frame to blame for the long task on the whole: this is refered to as "minimal frame attribution" and is captured in the name field
  • Pointing to the type of work involved in the task, and its associated frame context: this is captured in TaskAttributionTiming objects in the attribution field of PerformanceLongTaskTiming

Therefore, name and attribution fields on PerformanceLongTaskTiming together paint the picture for where the blame rests for a long task.

The security model of the web means that sometimes a long task will happen in an iframe that is unreachable from the observing frame. For instance, a long task might happen in a deeply nested iframe that is different from my origin. Or similarly, I might be an iframe doubly embedded in a document, and a long task will happen in the top-level browsing context. In the web security model, I can know from which direction the issue came, one of my ancestors or descendants, but to preserve the frame origin model, we must be careful about pointing to the specific container or frame.

Currently the TaskAttributionTiming entry in attribution is populated with "script" work (in the future layout, style etc will be added). The container or frame implicated in attribution should match up with the name as follows:

value of name frame implicated in attribution
self empty
same-origin-ancestor same-origin culprit frame
same-origin-descendant same-origin culprit frame
same-origin same-origin culprit frame
cross-origin-ancestor empty
cross-origin-descendant first cross-origin child frame between my own frame and culprit frame
cross-origin-unreachable empty
multiple-contexts empty
unknown empty

Privacy & Security

Long Tasks API surfaces long tasks greater than a threshold (50ms) to developers via Javascript (Performance Observer API). It includes origin-safe attribution information about the source of the long task. There is a 50ms threshold for long tasks. Together this provides adequate protection against security attacks against browser.

However, privacy related attacks are possible, while the API doesn’t introduce any new privacy attacks, it could expedite existing privacy attacks. If this were to become an concern, additional mitigations can be implemented to address this such as dropping "culprit" after a per-target origin threshold is exceeded, or limiting to 10 origins per minute etc.

Detailed Security & Privacy doc is here: https://docs.google.com/document/d/1tIMI1gau_q6X5EBnjDNiFS5NWV9cpYJ5KKA7xPd3VB8/edit#

V2 API Sketch

See: https://docs.google.com/document/d/125d69JAC7nyx-Ob0a9Z31d1uHUGu4myYQ3os9EnGfdU/edit

Alternatives Considered

Why not just show sub-tasks vs. top-level tasks with attribution?

This API will show toplevel long tasks along with attribution for specific sub-tasks which were problematic. For instance, within a 50ms toplevel task, sub-tasks such as a 20ms script execution or a 30ms style & layout update -- will be attributed. This raises the question -- why show the toplevel task at all? Why not only show long sub-tasks such as script, style & layout etc that are directly actionable by the user? The top level task may contain some un-attributable segments such as browser work eg. GC or browser events etc.

The rationale here is that showing the toplevel task is good for web developers, even though they will actively consume the actionable sub-tasks such as long scripts and act on them. Over time the sub-task attribution will keep expanding, making more of the long task actionable. Showing the top-level task gives developers a direct indication of main thread busy-ness, and since this directly impacts the user experience, it is appropriate for them to know about it as a problem signal -- even if they cannot have complete visibility or full actionability for the entire length of the long task. In many cases the developers may be able to repro in lab or locally and glean additional insights and get to the root cause. Long tasks provide context to long sub-tasks, for instance, a 20ms style and layout or a 25ms script execution may not be terrible by themselves, but if they happen consecutively (eg. script started from rAF) and cause a long 50ms task, then this is a problem for user responsiveness.