Logging every error frame is insane (hundreds of millions of logs).
We should do a few things:
- on change logging of the most recent error (maybe, although that could deteriorate to SPAM if there's multiple issues)
- rate limit and give periodic summaries
I think maybe we should do periodic summary reports if there are recent error frames, and leave logging the errors to ... the logger.
Logging every error frame is insane (hundreds of millions of logs).
We should do a few things:
I think maybe we should do periodic summary reports if there are recent error frames, and leave logging the errors to ... the logger.