What is rate-limiting configuration on getoutline.com?

[jledentu]

Hi,

Our organization uses the Outline API to retrieve some pages and generate static HTML content. It was working perfectly until today, when we met 429 (Rate limit exceeded) errors.

I'd like to know how Rate Limiting is configured on getoutline.com. I was not able to find the information in the API documentation. I'm curious because I don't think we call the API so much.

Here are the headers I received:

Retry-After: 59.693
RateLimit-Limit: 5
RateLimit-Remaining: 0
RateLimit-Reset: Wed Jan 31 2024 09:42:41 GMT+0000 (Coordinated Universal Time)

So the limit would be 5 calls by minute? Seems very low ??

Thanks!

[tommoor]

The rate limit depends on the API endpoint, which endpoint are you hitting limits on?

[jledentu]

POST /documents.export

https://www.getoutline.com/developers#tag/Documents/paths/~1documents.export/post

[tommoor]

https://github.com/outline/outline/blob/main/server/routes/api/documents/documents.ts#L519-L520

The limit is 5/minute as this endpoint is computationally expensive (PDF generation is performed through this endpoint). In hindsight it might be nice to offer a higher limit to HTML exports, but that would require some refactoring to only apply to the rate limit depending on input parameters.

[jledentu]

Thanks for the explanation, even if I don't understand why I didn't reach the rate limit until now. Our tooling is suddenly broken. So is there any way to retrieve the markdown content of a set (I mean more than 5) of documents via API ?

[tommoor]

You can either:

• Look at collections.export for the equivalent method on an entire collection, this is async though – so more complex to get the result.
• Update your code to respect the rate limit – there is a Retry-After header which gives you the number of ms you need to wait until next request (easiest).
• We can also adjust the rate limit here in the API to be higher for non-PDF exports, but it might take a day or two given the aforementioned refactor that would be required.

[jledentu]

Thanks @tommoor, a higher limit would be nice, for now I will try the options you exposed. Thanks again for your time.

[tommoor]

Limit is upped to 25/minute now, you should still handle the rate limit headers though 😉

[jledentu]

@tommoor Thank you so much! I now handle the rate limit headers. However, the new limit is less strict so the processing will stay fast.