Sending large number of statuses/updates


We have several thousand users and many of them schedule tweets to be sent on the hour. We’re managing to get them sent, but as membership grows, we see some potential issues.

The rest api seems to be very quick, but on some hours of the day, the time to send all of the updates is taking up to 5-8 minutes to post them all. (using two background processes to get it done)

All things being equal, as our membership grows, so will the time it takes to post all the updates. IE - if membership doubles then hourly scheduled tweets will take 10-20 minutes (or there about), and so on.

Curious how some of the big boys (say HootSuite as an example) manage to get several hundred thousand tweets posted “on time” - or do they?

Any suggestions or best practices to do this?


If the issue is servers, than it’s more of a parallelization issue. At that point, you’re looking at worker/queue infras, and things like Task Queues in Google Cloud.

Regarding rate limits (which I don’t believe is the issue), ensuring you have user auth should mean you’ll never see problems on the twitter front.


This topic was automatically closed 36 hours after the last reply. New replies are no longer allowed.