Control.set :telemetry_request_purge_job_wait, 1
Moved from deleting 500 rows every 2 seconds to every second. Easy to see on the uptick on @PlanetScale graphs.
Controlling things at runtime is the way.
| Website | https://johnnunemaker.com |
John Nunemaker
@jnunemaker@ruby.social
- 302 Followers
- 111 Following
- 599 Posts
Deploying without feature flags is like coding without version control.
Continuously purge data from mysql via good job in a safe way. Controlled by feature flag and runtime controls for how many records to delete in a batch and how long to wait to run the job again (to control replication lag).
Easy to turn off, slow down or speed up at runtime!
Flipper 1.3.5 is out. Mostly just some effort to make the active record and redis adapters Just Work™ more often than they were.
https://blog.flippercloud.io/flipper-1-3-4-and-1-3-5/
Enjoy!
We shipped out tshirts to some OG fireside customers. We did it to say thanks. But caring is always good business. Today, I got an email from one of the customers who just moved another podcast to us. They said the tshirt reminded them they've been wanting to migrate it. 🔥🔥🔥
Today I saw someone with a neovim license plate at wind cave national park and said hi. Programmers are everywhere! 😊
Just ran into someone in a datadog shirt in the badlands. Of course I said hi. She said her brother works there. 😊
So many goodies in rails of extras — particularly measure_queries. A lot of overlap with pghero.
Anthropic is a real one for this. Check out their console message if you are bored. I'm sure they copied someone else but this is a brilliant idea.
