502 errors between postgres and app for large queries

Hi there,

I have an axum application running, connected to a postgres database, both on fly.

I am noticing that certain API calls, especially the ones querying larger date ranges, are returning 502 in the application. Here are some snippets from the logs:

2023-07-16T20:28:49.970 app[e286376be17e86] jnb [info] 2023-07-16T20:28:49.969Z INFO [sqlx_core::pool::inner] ping on idle connection returned error: error communicating with database: connection aborted 
2023-07-16T20:21:16.872 proxy[e286376be17e86] jnb [error] could not make HTTP request to instance: connection closed before message completed

What I can glean from these logs is that the application is struggling to maintain a connection to postgres. And that the connection between app and db is being closed prematurely.

I have a suspicion that the TCP connection is simply timing out before the db finishes the query, as this issue only happens when I query a large date range in the application. What is the timout limit for inter-app communication? Is it configurable?

Beyond that, any feedback or advice would be much appreicated. Please let me know if you require more information from my side. Thanks!

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.