Replication
Copying data across nodes for redundancy and read-scaling — leader-follower vs multi-leader.
Why we need this / what value this brings
Protects against data loss on primary failure and lets you offload read traffic from the primary.
When to use this
Once read load noticeably affects write performance, or availability requirements justify a hot standby.
How to use or implement this
Start with a single read replica for reporting/analytics queries before investing in multi-region or multi-leader setups.
Research questions
- Does Localz's Postgres setup use a read replica yet, and would it help the storefront browsing load?
Empty folder — drop notes, links, and findings here as you research.