Francesco Nigro
Francesco Nigro
Currently Techempower benchmarks can/won't delete docker containers depending if debug/benchmark mode is used to run it: it won't help collecting tracing/profiling and/or other stats collected by containers on test stop....
This is an alternative to #38737 with no need of cutoff values and always performing fast-path unlocked checks, if possible, and lazy allocation of a `ReentrantLock`. In order to reduce...
In the OpenTelemetry benchmark (run by @barreiro and @brunobat ) I've spotted some memory and cpu waste around `VertxUtil` related how we lookup for known HTTP headers and how we...
This could hit a performance issue (see https://stackoverflow.com/questions/44988841/force-tableswitch-instead-of-lookupswitch), because the default `-XX:MinJumpTableSize` is 10 and beyond it, the JIT would produce a jump table, instead of having a sort-of binary...
This is addressing 3 performance problems: - local/non-local context data's operations shouldn't use costy `synchronized` on existing maps (now that biased locking is gone) - local/non-local context data's lookup/remove doesn't...
While inspecting few benchmark this behaviour popup in the HTTP 2 server response handling:  It shows that we do perform few validations around header names, which could be improved...
As reported (and fixed) at https://github.com/netty/netty/pull/12806, only `ChannelDuplexHandler`s in netty have been fixed to solve the infamous type pollution issue: by running an hello-world like test with HTTP 2 in...
From https://github.com/eclipse-vertx/vert.x/blob/6a8aefb6ec84278774d2c03fdae83fdf1deafded/src/main/java/io/vertx/core/impl/VertxImpl.java#L104 it appears that Vertx, if not configured to do things differently, just set the Netty leak detection level to be DISABLED. Together with https://github.com/netty/netty/pull/13459 this can make such...
It would be great an option to enforce acquisition of connections just from the same context, for maximum performance or been parked if not found any; together with some telemetry...
HI @nitsanw ! Now, for convenience we create the unpadded version of the queues out of the Unsafe-based ones, but actually, if users doesn't care about maximum performance by preventing...