Comment by levkk
16 hours ago
Not quite. The performance gain is to bring those features to Postgres!
Edit:
Performance gains are from having the ability to load balance reads (horizontal scaling for read queries) and scale out writes (with sharding). Once instance bottleneck in Postgres has many faces:
1. Behind schedule vacuums because of too many dead tuples (too many writes)
2. The WALWriter is single-threaded and IO-bound - Postgres can only do about 200-300MB/sec in writes per instance (real prod numbers on EC2 with NVMes and ZFS, basically best case scenario).
3. Bulkheading: single primary is a single point of failure. With 12 primaries, if one fails, 91% of your customers don't notice.
The list goes on. Rust is just a side effect. We love it because it's fast and correct - the perfect match for a database product.
So to oversimplify, is the idea to bring an AWS Aurora-style storage mechanism natively to Postgres?
Aurora is one big database, isn't it? PgDog is just a proxy where you tell it which shard to access.
Yes, except it doesn't have any cross-dependencies on the same volume, so the uptime here should be higher.
Aurora has a completely different storage backend. PgDog is a front end proxy - each server in the cluster is still using standard Postgres right?
1 reply →
Oh thanks for clearing that up.
Sorry, out walking the dog (not a pun). I'll post more details in a few.