Health check logging: ElapsedMilliseconds in messages makes logs harder to process #21608
Description
Disclamer: I think logging the duration of health checks is great.
I also think it should be logged as debug only. The current code logs the duration in warnings and errors too - those who are important to monitor in production. In Splunk we group log messages before creating a daily report. When duration is the only differing factor our reports look like this:
Proposal: only include ElapsedMilliseconds in debug messages
And the link to the code