>Without those periodic full page images in the log, the storage layer would have to replay an infinitely long chain of small deltas to reconstruct a page for a read request. What was once a bounded O(checkpoint frequency) replay becomes an unbounded chain, leading to a spike in read latency and resource consumption.
I don't follow: read requests are not served from the WAL. They read the current state of the page from the buffer cache, where the page is updated after the change (FPI or not) is written to the WAL.
I don't follow: read requests are not served from the WAL. They read the current state of the page from the buffer cache, where the page is updated after the change (FPI or not) is written to the WAL.
This appears to only have any effect with datalake style installs, where storage is separate from compute.
Not going to have any effect on those small postgres installs for that generic one off app.