4 ms·
An interesting article about how Twitch manages their Postgres cluster of around 100 machines. Also their approach to use Aws autoscaling to scale read replica
by pritambarhate 10y ago
An interesting article about how Twitch manages their Postgres cluster of around 100 machines.
Also their approach to use Aws autoscaling to scale read replicas is quite new and novel to me.
- merb 10y agowell I wonder how much man power they spend into their whole cluster. I mean look at how much setting tweaking / load balacing / auto scale / haproxy / pgbouncer / upgrade.. For the upgrade part I hope https://2ndquadrant.com/de/resources/pglogical/pglogical-docs/ https://2ndquadrant.com/de/resources/pglogical/pglogical-doc... will get better and better.
- ben_jones 10y agoAlso the relative smorgasbord of DBMS that they use aside from postgres. I definitely appreciate their transparency though it's hard to miss the "And we're hiring" plug at the end of it.
- merb 10y ago> it's hard to miss the "And we're hiring" plug at the end of it. well I missed it. O.O