On July 20, 2021, between 15:10 and 17:30 UTC, the Heroku Runtime team became aware of an issue that would fail to start new Performance-M and Performance-L class dynos in a subset of our fleet. We sincerely apologize for the negative effects our customers experienced.
Who was affected?
Cu…
Resolved
Dynos can now be started.
Monitoring
The fix has been completely deployed. Apps that had temporarily dyno types changed can now safely move back to the original types.
Please file a support ticket if you shall still see the impact.
Update
We have discovered Performance-M dynos have been affected in addition to Performance-L dynos. Engineers are rolling back the bad deployment. The roll back is almost complete.
Issue
Engineers are looking into Performance-L dynos being unable to start with errors like:
Error waiting for network: Resource temporarily unavailable
There are reports that changing dyno size to Standard-2X made it possible to start dynos.
Investigating
Engineers are looking into reports of dynos being unable to start. App logs may show errors like:
Error waiting for network: Resource temporarily unavailable