Feature Requests

If log type of Error appears more than X times in time interval of Y, I'd like to get a notification on Email/Slack etc

I would like to be able to use any monitory service I want for my managed services like postgres. This usually requires allowing this third-party to scrape opentelemetry metrics from your services. Easily doable when you control the runtime, but not so much when it is a managed service like Postgres or a load balancer. Example of very important metrics: Rate of status code (especially 5xx and 4xx) RAM/CPU usage in the database

In the log viewer, Render infers log levels based on certain keywords like "error". I find this undesirable and just want the original log levels to be respected. For example, I'm logging an expected error at the info level with a message like "Expected error..." and Render's log viewer captures it as "level:error" even though I logged it at the info level. I can see this inference being useful for logging systems without levels, but if the logging system has levels, just respect those and don't change them based on inference.

Currently Logstreams only support TCP endpoints. However, some services like Grafana Loki only expose HTTPS endpoints. Allowing users to choose between TCP or HTTPS endpoints would make more logging services compatible with Render.

Support for connecting to self-hosted and/or cloud version of Loki via LogStreams https://github.com/grafana/loki

Hi render friends! I just learned about https://render.com/docs/static-sites#dependency-installation while troubleshooting a static site deployment. I was thinking: right before render pulls down the dependencies, a log line saying something like: "Automatically pulling dependencies via npm, see https://render.com/docs/static-sites#dependency-installation for details" would have saved me some head scratching. I definitely don't mind that it happens automatically, just thinking a bit more of a heads up would help!

A couple of things that would improve the logs being sent to Datadog: * "syslog.appname" and "syslog.hostname" should be swapped. "syslog.appname" maps to the "service" attribute in Datadog, which shouldn't changed based on the number of instances being run. "syslog.hostname" maps to the "host" attribute in Datadog, which should change based on the number of instances being run. * Some logs that should be "info" are showing up as "alert". For example, Sidekiq logs about a job completing include "lvl: 'INFO'" in the log, but are showing up in Datadog as "alert".

I am sending all Render logs to DataDog and it's becoming a huge bill. I would like to selectively turn off logging for services and within one service to be able to disable certain kind of logs (like request logging) or to limit them (only log X% requests).

Log is too slow during active debugs actions. Probably is a good idea to improve the performance and print the log every 100 ms when someone is in Live mode and not lost a great part of chron-jobs and functions feedback (sometime it happen)

I can search application and request logs using the built-in log explorer in the Dashboard. However, I would like the ability to export the logs without needing an external provider (e.g., log streams). I want to export a subset of filtered logs to a static file for offline analysis. There should be two export formats: Plain text file: pretty much what I see in the UI CSV file: structured data, where there are columns for log level, status code, HTTP method, request ID, etc.. Timestamps in exported logs should also be in UTC, not localized.