Maximizing CockroachDB Performance: Our Journey to 1M QPS
1–5 of 5 posts
Re: Maximizing CockroachDB Performance: Our Journey to 1M QPS
#2Learn how AuthZed scaled SpiceDB on CockroachDB to 1 million authorization events per second with our now open-sourced advanced connection pooler, crdbpool. Discover the challenges we faced and how we solved them in their journey to maximizing CockroachDB performance.
Re: Maximizing CockroachDB Performance: Our Journey to 1M QPS
#3Learn how AuthZed scaled SpiceDB on CockroachDB to 1 million authorization events per second with our now open-sourced advanced connection pooler, crdbpool. Discover the challenges we faced and how we solved them in their journey to maximizing CockroachDB performance.
what is the size/hardware spec of the cluster?
Re: Maximizing CockroachDB Performance: Our Journey to 1M QPS
#4Learn how AuthZed scaled SpiceDB on CockroachDB to 1 million authorization events per second with our now open-sourced advanced connection pooler, crdbpool. Discover the challenges we faced and how we solved them in their journey to maximizing CockroachDB performance.
what is the size/hardware spec of the cluster?
We're saving some of that for a future blog post - we tested many different configurations and scales (of both SpiceDB and CRDB)
Re: Maximizing CockroachDB Performance: Our Journey to 1M QPS
#5Thanks for making the retry logic as it's own library! I would have thought crdb would have it's own logic for when a node is draining but nice to be able to handle that on the client side.