codelipenghui commented on code in PR #24457: URL: https://github.com/apache/pulsar/pull/24457#discussion_r2167269561
########## pulsar-common/src/main/java/org/apache/pulsar/common/naming/NameCache.java: ########## @@ -0,0 +1,156 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, + * software distributed under the License is distributed on an + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + * KIND, either express or implied. See the License for the + * specific language governing permissions and limitations + * under the License. + */ +package org.apache.pulsar.common.naming; + +import com.google.common.collect.Interner; +import com.google.common.collect.Interners; +import java.lang.ref.ReferenceQueue; +import java.lang.ref.SoftReference; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ConcurrentMap; +import java.util.concurrent.atomic.AtomicBoolean; +import java.util.concurrent.atomic.AtomicLong; +import org.apache.pulsar.common.util.StringInterner; + +/** + * A cache for TopicName and NamespaceName instances that allows deduplication and efficient memory usage. + * It uses soft references to allow garbage collection of unused instances under heavy memory pressure. + * This cache uses ConcurrentHashMap for lookups for performance over Guava Cache and Caffeine Cache + * since there was a concern in https://github.com/apache/pulsar/pull/23052 about high CPU usage for TopicName lookups. + */ +abstract class NameCache<V> { + // Cache instances using ConcurrentHashMap and SoftReference to allow garbage collection to clear unreferenced + // entries when heap memory is running low. + private final ConcurrentMap<String, SoftReferenceValue> cache = new ConcurrentHashMap<>(); + // Reference queue to hold cleared soft references, which will be used to remove entries from the cache. + private final ReferenceQueue<? super V> referenceQueue = new ReferenceQueue<>(); Review Comment: > Only unused TopicNames will get collected. If the broker or the application holds a strong reference, it won't get removed from the cache. What is the concern? My concern is unpredictability since the decision made by JVM is based on heuristics and is influenced by multiple factors, different JVM implemetation, different GC algorithms, different versions. We don't want to see the topic names cache get removed and reloaded frequently, it's not data. Especially when you have a tiny broker (200MB heap memory and 100m CPU assigned). For example, the topic names cache might take 20MB in this case, if the topic names cache gets removed by the JVM GC frequently, the topic name cache clear and reload might exhaust all the CPU resources. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@pulsar.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
