[improve][broker] Make Consumer#equals more effective

[Shawyeok]

Motivation

It's from a production case in my org, one of our consumer application constantly create consumer instance after increase topic partition due to old pulsar-client bug. The application got OOM eventually and broker publish latency goes more than 30s, so many publish requests timeout, broker latency go back normal after application restart.

It could be reproduce below:

pulsar-admin namespaces create -b 128 public/perf
pa namespaces set-max-consumers-per-topic -c 0 public/perf
pa namespaces set-max-consumers-per-subscription -c 0 public/perf
pulsar-perf produce -r 1000 persistent://public/perf/perf
pulsar-perf consume -st Shared -n 100000 persistent://public/perf/perf

After stop consume program, pulsar-io thread will be block for a while, publish latency will be very high (35s+ in my case).

It's a cpu profile during consumer stops:

• profile.100kconsume1k.stoping.122601.html.zip

The time complexity of CopyOnWriteList.remove is O(n), so Consumer#equals method is a hot method.
https://github.com/apache/pulsar/blob/master/pulsar-broker/src/main/java/org/apache/pulsar/broker/service/persistent/PersistentDispatcherMultipleConsumers.java#L192

Modifications

Flip && expression

Verifying this change

• Make sure that the change passes the CI checks.

This change is a trivial rework / code cleanup without any test coverage.

Documentation

• doc-not-needed

Matching PR in forked repository

PR in forked repository: Shawyeok#5

[hangc0276]

Nice catch!

[tisonkun]

Thank you! Merging...