Skip to main content

Request Scheduler

From-v2.2

The request scheduler enables an application to "issue" a large number of requests without flooding the browser's limited request queue.

A getPriority callback is called on all outstanding requests whenever a slot frees up, allowing the application to reprioritize or even cancel "issued" requests if the application state has changed.

Note: The request scheduler does not actually issue requests, it just lets apps know when the request can be issued without overwhelming the connection and the server.

A primary use case is to let the app reprioritize or cancel requests if circumstances change before the request can be scheduled.

Usage​

To schedule a request so that it can be issued at a time when it can be immediately processed.

const URL = '...';
const requestToken = await requestScheduler.scheduleRequest(URL);
if (requestToken) {
  await fetch(URL);
  requestToken.done(); // NOTE: **must** be called for the next request in queue to resolve
}

Methods​

constructor(options?: object)​

  • id?: string;
  • throttleRequests?: boolean; — Whether to throttle requests.
  • maxRequests?: number; — Maximum concurrent requests to be opened by this scheduler. Additional requests are queued until an open request has completed.
  • debounceTime?: number; — Queue requests until no new requests have been added for at least debounceTime milliseconds.

When throttleRequests is disabled, maxRequeusts and debounceTime options have no effect.

scheduleRequest(handle: any, getPriority?: () => number): Promise<{done: () => any)}>​

Called by an application that wants to issue a request, without having it deeply queued by the browser

When the returned promise resolved, it is OK for the application to issue a request. The promise resolves to an object that contains a done method. When the application's request has completed (or failed), the application must call the done function

Parameters

  • handle an arbitrary handle to identify the request, e.g. a URL string
  • getPriority will be called when request "slots" open up, allowing the caller to update priority or cancel the request Highest priority executes first, priority < 0 cancels the request

Returns a promise that

  • resolves to an object (with a done field) when the request can be issued without queueing. The application should issue the request and call done() when completed.
  • resolves to null if the request has been cancelled (by the callback return < 0). In this case the application should not issue the request.

About Request Priorities​

The getPriority callback controls priority of requests and also cancellation of outstanding requests.