Thomas Skjølberg

@bdemers I agree. The remote JWKs should be kept up-to-date; refreshes triggered either by time or by unknown keys. JWKs no longer on the list should be banned. Thus changes...

@lhazlewood there is not much of a difference in performance by it self, but gotta start somewhere; doing unecessary stuff on the critical path should be avoided. I guess creating...

@DanFTRX well something is better than nothing, and this PR also simplifies the logic a bit, as a bonus. But I think if I wrote this again, I'd simply extend...

As a side note: The SLF4J 1.8 beta seems to work with Logback 1.2.3, if you return `StaticLoggerBinder.getSingleton().getLoggerFactory()` in your custom `SLF4JServiceProvider` `getLoggerFactory(..)` method.

@juncevich out of interest, which web server are you running?

I wonder if gRPC logging might be better served with a gRPC-centric module. At least thats how we're doing it internally (emulating the Logbook output format). Also the HTTP 2...

So would it make sense to first go with a JMH style benchmarking and then perhaps later move on to a full spring boot example app + load testing tool...

Note on JsonHttpLogFormatter - it slows it down first calling prepare(..) then format(..) (in StructuredHttpLogFormatter), that should really be a single step for converting right to JSON.

Well at least wasteful. A JsonWriter or (even faster) a StringBuilder is preferable.

These jackson objects are a bit heavy, so my guess is that this will be in the 10 percent range improvement. Filtering the payloads will be the most demanding part...