[GitHub] [ozone] jojochuang commented on a change in pull request #2919: HDDS-6086. Compute MD5MD5CRC file checksum using chunk checksums from DataNodes

[GitBox <gi...@apache.org>]

jojochuang commented on a change in pull request #2919:
URL: https://github.com/apache/ozone/pull/2919#discussion_r778625639



##########
File path: hadoop-ozone/client/src/main/java/org/apache/hadoop/ozone/client/ReplicatedFileChecksumHelper.java
##########
@@ -0,0 +1,181 @@
+/*
+ * 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.hadoop.ozone.client;
+
+import org.apache.hadoop.fs.PathIOException;
+import org.apache.hadoop.hdds.client.BlockID;
+import org.apache.hadoop.hdds.client.ReplicationConfig;
+import org.apache.hadoop.hdds.client.StandaloneReplicationConfig;
+import org.apache.hadoop.hdds.protocol.datanode.proto.ContainerProtos;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.XceiverClientSpi;
+import org.apache.hadoop.hdds.scm.pipeline.Pipeline;
+import org.apache.hadoop.hdds.scm.storage.ContainerProtocolCalls;
+import org.apache.hadoop.hdds.security.token.OzoneBlockTokenIdentifier;
+import org.apache.hadoop.io.MD5Hash;
+import org.apache.hadoop.ozone.client.rpc.RpcClient;
+import org.apache.hadoop.ozone.om.helpers.OmKeyLocationInfo;
+import org.apache.hadoop.security.token.Token;
+
+import java.io.IOException;
+import java.util.List;
+
+/**
+ * The helper class to compute file checksum for replicated files.
+ */
+public class ReplicatedFileChecksumHelper extends BaseFileChecksumHelper {
+  private int blockIdx;
+
+  ReplicatedFileChecksumHelper(
+      OzoneVolume volume, OzoneBucket bucket, String keyName, long length,
+      RpcClient rpcClient) throws IOException {
+    super(volume, bucket, keyName, length, rpcClient);
+  }
+
+  @Override
+  protected void checksumBlocks() throws IOException {
+    long currentLength = 0;
+    for (blockIdx = 0;
+         blockIdx < getKeyLocationInfoList().size() && getRemaining() >= 0;
+         blockIdx++) {
+      OmKeyLocationInfo keyLocationInfo =
+          getKeyLocationInfoList().get(blockIdx);
+      currentLength += keyLocationInfo.getLength();
+      if (currentLength > getLength()) {
+        return;
+      }
+
+      if (!checksumBlock(keyLocationInfo)) {
+        throw new PathIOException(
+            getSrc(), "Fail to get block MD5 for " + keyLocationInfo);
+      }
+    }
+  }
+
+  /**
+   * Return true when sounds good to continue or retry, false when severe
+   * condition or totally failed.
+   */
+  private boolean checksumBlock(OmKeyLocationInfo keyLocationInfo)
+      throws IOException {
+
+    long blockNumBytes = keyLocationInfo.getLength();
+
+    if (getRemaining() < blockNumBytes) {
+      blockNumBytes = getRemaining();
+    }
+    setRemaining(getRemaining() - blockNumBytes);
+    // for each block, send request
+    List<ContainerProtos.ChunkInfo> chunkInfos =
+        getChunkInfos(keyLocationInfo);
+    ContainerProtos.ChecksumData checksumData =
+        chunkInfos.get(0).getChecksumData();
+    int bytesPerChecksum = checksumData.getBytesPerChecksum();
+    setBytesPerCRC(bytesPerChecksum);
+
+    byte[] blockChecksum = getBlockChecksumFromChunkChecksums(
+        keyLocationInfo, chunkInfos);
+    String blockChecksumForDebug = populateBlockChecksumBuf(blockChecksum);
+
+    LOG.debug("got reply from pipeline {} for block {}: blockChecksum={}, " +
+            "blockChecksumType={}",
+        keyLocationInfo.getPipeline(), keyLocationInfo.getBlockID(),
+        blockChecksumForDebug, checksumData.getType());
+    return true;
+  }
+
+  // copied from BlockInputStream
+  /**
+   * Send RPC call to get the block info from the container.
+   * @return List of chunks in this block.
+   */
+  protected List<ContainerProtos.ChunkInfo> getChunkInfos(
+      OmKeyLocationInfo keyLocationInfo) throws IOException {
+    // irrespective of the container state, we will always read via Standalone
+    // protocol.
+    Token<OzoneBlockTokenIdentifier> token = keyLocationInfo.getToken();
+    Pipeline pipeline = keyLocationInfo.getPipeline();
+    BlockID blockID = keyLocationInfo.getBlockID();
+    if (pipeline.getType() != HddsProtos.ReplicationType.STAND_ALONE) {
+      pipeline = Pipeline.newBuilder(pipeline)
+          .setReplicationConfig(new StandaloneReplicationConfig(
+              ReplicationConfig
+                  .getLegacyFactor(pipeline.getReplicationConfig())))
+          .build();
+    }
+
+    boolean success = false;
+    List<ContainerProtos.ChunkInfo> chunks;
+    XceiverClientSpi xceiverClientSpi = null;
+    try {

Review comment:
       Ok. Good catch. We should dedup it. However, looking at the code, HddsClientUtils is not a good abstraction. Instead, ContainerOperationClient looks to be a better place for the code.
   
   In fact, the BlockInputStream should not in the business of managing the XceiverClient connection at all. I would vow for rewriting BlockInputStream to use ContainerOperationClient APIs to manage the connections. It's unrelated to this PR so I'll raise another jira for that.




-- 
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-unsubscribe@ozone.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org