4 ms·
Your statement intrigued me, so I fired up the ol' googles and started looking to see if anyone had tried this. And within the first page of results I found a c
by drewbug01 6y ago
Your statement intrigued me, so I fired up the ol' googles and started looking to see if anyone had tried this. And within the first page of results I found a comment from you a few months ago saying the same thing! :)
This seems really interesting - at least for debugging (I worry that it would tank performance under load). Have you considered trying to work on it? My googling suggest that you seem rather interested in the idea! The postgres community is overall really welcoming to contributions (as is the OpenTelemetry community, hint hint).
- candiddevmike 6y agoI keep posting it on HN hoping a Postgres dev hears my plea :) I've never programmed in C for anything serious, so I'm not sure where I'd even start. I _think_, based on my limited knowledge of postgres extensions, you'd have to bake the jaeger sampling into PG proper--I don't think extensions can intercept/inspect triggers.
- sitkack 6y agoI think the solution would be to add triggers to the dtrace probes. https://www.postgresql.org/docs/current/dynamic-trace.html https://www.postgresql.org/docs/current/dynamic-trace.html