neon
neon copied to clipboard
Add context for `Timed out while waiting for WAL` error
Distinguish between 4 cases: a) We have a walreceiver and we reached end of stream b) We're actively processing wal c) There's no walreceiver d) Requested LSN is ahead of commit LSN
Or possibly measure and report breakdown of what all the time was spent on. Ideally we'd use tracing infrastructure but we can improvise until then.
Hi, I opened a PR to address this. It is not very elegant solution, but I guess as the ticket says, this simply adds a bit of context whenever the timeout occurs.
I also think the wording in the context can be slightly improved, but anyway, please have a look if I can improve something there.