profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/hagbard/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.
David Beaumont hagbard @google , @googlei18n Zurich

google/flogger 1317

A Fluent Logging API for Java

hagbard/asmble 0

Compile WebAssembly to JVM and other WASM tools

hagbard/cldr 0

The new home of the Unicode Common Locale Data Repository

hagbard/icu 0

The new home of the ICU project source code.

pull request commentgoogle/flogger

Log4j2 backend uses ScopedLoggingContext

It seems to work, but I wonder if this approach is sufficient.

Thanks! I'm a little concerned about the cost talking to the log manager during each log statement (some methods in there seem to talk about classloaders), so maybe we can cache whether it has an appendable (do we expect this to change much). I guess the risk with that is that during startup we'll appear to see no appenders before the config code is called.

I'm also suspicious that we'd need to look at the appenders of the parent loggers too (I don't know how Log4J does this, but JDK logger defers to the parent logger's handlers). It seems odd that log4J would expect people to explicitly add appenders to every logger in the hierarchy (rather than just the one at the top).

However since I don't really know much about this, I'll defer to code if it's reasonably well tested. I think we can probably take things from here (you've done a lot of work already on this and I think the log4j part of the story is on us to look at). I mean if you want to keep looking at it, feel free, but don't feel you're expected to.

sekaiser

comment created time in 2 months