mastodon: add proper rate limiting #21

New Issue

Open

opened 2023-03-20 10:03:26 +00:00 by missytake · 0 comments

missytake commented

2023-03-20 10:03:26 +00:00

Owner

Right now, we only send one poll per minute.

https://mastodonpy.readthedocs.io/en/stable/01_general.html#rate-limiting mentions that the default is 300 requests in 5 minutes, that would allow polling once per second instead of once per minute. But we don't know in advance how often we're gonna post, and the rate limit can be non-default.

Using the pace rate limit option is probably not possible with async? What we could use instead is the throw option to throw a MastodonRateLimitError if we run against the rate limit, request https://mastodonpy.readthedocs.io/en/stable/01_general.html#rate-limiting, and decide upon that for how long we do an await sleep().

Or even better, use https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_remaining and https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_reset to guess how long we should sleep between requests, and if we hit a MastodonRateLimitError, we wait until https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_reset.

Right now, we only send one poll per minute. https://mastodonpy.readthedocs.io/en/stable/01_general.html#rate-limiting mentions that the default is 300 requests in 5 minutes, that would allow polling once per second instead of once per minute. But we don't know in advance how often we're gonna post, and the rate limit can be non-default. Using the `pace` rate limit option is probably not possible with async? What we could use instead is the `throw` option to throw a `MastodonRateLimitError` if we run against the rate limit, request https://mastodonpy.readthedocs.io/en/stable/01_general.html#rate-limiting, and decide upon that for how long we do an `await sleep()`. Or even better, use https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_remaining and https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_reset to guess how long we should sleep between requests, and if we hit a `MastodonRateLimitError`, we wait until https://mastodonpy.readthedocs.io/en/stable/01_general.html#mastodon.Mastodon.ratelimit_reset.