Luis

Results 5 issues of Luis

When using "EXPLAIN (ANALYZE, TIMING)", IO timings can be confusing when measuring parallel nodes. ![BDjxgWEcof](https://user-images.githubusercontent.com/26086198/98027017-c0417480-1dea-11eb-966a-ff16244d5de2.png) Example: https://explain.dalibo.com/plan/avr0 What do you think of showing the maximum time between all wokers on...

question

``` WARNING: DATA RACE Read at 0x00c000ed1e58 by goroutine 46: github.com/asynkron/protoactor-go/cluster.(*Gossiper).GetActorCount() /Users/lrweck/go/pkg/mod/github.com/asynkron/[email protected]/cluster/gossiper.go:344 +0x254 github.com/asynkron/protoactor-go/cluster.(*Gossiper).gossipLoop() /Users/lrweck/go/pkg/mod/github.com/asynkron/[email protected]/cluster/gossiper.go:331 +0x1ec github.com/asynkron/protoactor-go/cluster.(*Gossiper).StartGossiping.gowrap1() /Users/lrweck/go/pkg/mod/github.com/asynkron/[email protected]/cluster/gossiper.go:291 +0x34 Previous write at 0x00c000ed1e58 by goroutine 11463: ??() -:0 +0x104f034a8 sync/atomic.AddInt32()...

bug report

reflect causes a lot of memory allocations and uses more CPU time. maps.Equal does not cause any allocations and is about 80% faster ``` BenchmarkMapComparing/reflect.DeepEqual/10-8 1000000 1012 ns/op 320 B/op...

**Describe the bug** ``` WARNING: DATA RACE Write at 0x00c006720d40 by goroutine 11743: github.com/asynkron/protoactor-go/actor.(*actorContextExtras).killReceiveTimeoutTimer() /codefresh/volume/.cache/go-path/pkg/mod/github.com/asynkron/[email protected]/actor/actor_context.go:81 +0x88 github.com/asynkron/protoactor-go/actor.(*actorContext).CancelReceiveTimeout() /codefresh/volume/.cache/go-path/pkg/mod/github.com/asynkron/[email protected]/actor/actor_context.go:246 +0xf6 github.com/asynkron/protoactor-go/actor.(*actorContext).receiveTimeoutHandler() /codefresh/volume/.cache/go-path/pkg/mod/github.com/asynkron/[email protected]/actor/actor_context.go:252 +0x8f github.com/asynkron/protoactor-go/actor.(*actorContext).receiveTimeoutHandler-fm() :1 +0x33 Previous read at 0x00c006720d40 by...

I get consistent panics on the memberset call to remove members while running in cluster mode. It is not immediate, but after some time idling. I wasn't able to identify...