zhztheplayer commented on a change in pull request #7030: URL: https://github.com/apache/arrow/pull/7030#discussion_r440075959
########## File path: java/dataset/src/main/java/org/apache/arrow/memory/NativeUnderlingMemory.java ########## @@ -0,0 +1,65 @@ +/* + * 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.arrow.memory; + +import org.apache.arrow.dataset.jni.JniWrapper; + +/** + * AllocationManager implementation for Native allocated memory. + */ +public class NativeUnderlingMemory extends AllocationManager { Review comment: Hi @jacques-n, Thanks for your suggestion but I think the "weaknesses" of current implementation might be somehow overrated. Regarding the weaknesses, it's of course urgent for orc adaptor to improve its memory allocation process because all of the allocated memory by orc is being escaped from Java allocator. However we don't have the same problem here in Datasets, allocator is going to manage all the native memory. The only issue we may have is that OOM error might be thrown after a buffer is allocated rather than before we create it. But because OOM will be finally thrown and over-allocated memory will be finally released, the impact will be very very small. And I knew we can possibly change to allocate Java memory directly from C++ via JNI, but codes will become messy and performance will turn to worse. That is why I didn't propose such a solution in this PR. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
