uschindler commented on code in PR #11796: URL: https://github.com/apache/lucene/pull/11796#discussion_r983415631
########## lucene/misc/src/java/org/apache/lucene/misc/store/ByteWritesTrackingDirectoryWrapper.java: ########## @@ -0,0 +1,118 @@ +/* + * 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.lucene.misc.store; + +import java.io.IOException; +import java.util.concurrent.atomic.AtomicLong; +import org.apache.lucene.store.Directory; +import org.apache.lucene.store.FilterDirectory; +import org.apache.lucene.store.IOContext; +import org.apache.lucene.store.IndexOutput; + +/** {@link FilterDirectory} that tracks write amplification factor */ +public final class ByteWritesTrackingDirectoryWrapper extends FilterDirectory { + + private final AtomicLong flushedBytes = new AtomicLong(); + private final AtomicLong mergedBytes = new AtomicLong(); + private final AtomicLong realTimeFlushedBytes = new AtomicLong(); Review Comment: I would add a local plain `int` field (an int is enough here, as we only write small amounts and it fits better to the parameters passed to the methods) to the IndexOutput, increment it on every write completely unsychronized. After every write call a private method like `maybeUpdateGlobal()` which checks in the plain `int` field if it is larger than a certain value and then does something like `AtomicLong#addAndGet(localIntValue); localIntValue = 0;`. Of course if somebody is writing a large `byte[]` we can directly update the atomic long (to prevent overflows of the int). Nevertheless we should have a benchmark how much this affects writing. If the filter directory is just for debugging purposes, we can of course go with the full atomic implementation as we have now. Nobody takes care if you want to just watch the amplification factors while debugging your app and it gets slower because of this. -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
