A build-up of #Laravel jobs was causing one of our #BigCommerce apps to have poor UX.
A new release tackles this by:
- Prioritising user-initiated jobs
- Spreading cron jobs over the day
- Scaling #heroku dynos when queue size increases @hirefireio
https://t.co/ckPr2VChG6
@jamesarosen We've received similar feedback in the past (though, not 10-100x!). Not sure how to justify that kind of pricing tier. Something to think about. Thanks for your feedback, James! Definitely appreciate it.
@MaltheJorgensen Thanks!
Configuration is highly dependent on your application. Keeping timeout at 1 and using a low upscale/high downscale sensitivity value is a common approach. This allows you to respond to load increases faster, while scaling back slowly afterwards. Could help with flapping.
@sentwithlasers The open source variant has been archived as it's no longer being maintained, nor is it compatible with Heroku's current stack. The commercial version at https://t.co/VFwjV8cXGO is actively maintained.
@kev_the_dev It's not in the documentation. It's really straightforward when you're setting up a dyno managers. Within the same form you have the option to enable threshold notifications.
@Theo_Summe can you provide me with the name of your app? I’ll take a look. We’ve upgraded from API v2 to v3 years ago, so the brownout shouldn’t matter