[GitHub] [hbase] infraio commented on a change in pull request #832: HBASE-23298 Refactor LogRecoveredEditsOutputSink and BoundedLogWriter…

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

infraio commented on a change in pull request #832: HBASE-23298 Refactor LogRecoveredEditsOutputSink and BoundedLogWriter…
URL: https://github.com/apache/hbase/pull/832#discussion_r347177592
 
 

 ##########
 File path: hbase-server/src/main/java/org/apache/hadoop/hbase/wal/BaseRecoveredEditsOutputSink.java
 ##########
 @@ -0,0 +1,281 @@
+/**
+ * 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.hbase.wal;
+
+import static org.apache.hadoop.hbase.wal.WALSplitUtil.getCompletedRecoveredEditsFilePath;
+import static org.apache.hadoop.hbase.wal.WALSplitUtil.getRegionSplitEditsPath;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Map;
+import java.util.concurrent.ConcurrentHashMap;
+
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.FileSystem;
+import org.apache.hadoop.fs.Path;
+import org.apache.hadoop.hbase.Cell;
+import org.apache.hadoop.hbase.CellUtil;
+import org.apache.hadoop.hbase.HConstants;
+import org.apache.hadoop.hbase.log.HBaseMarkers;
+import org.apache.hadoop.hbase.util.Bytes;
+import org.apache.hadoop.hbase.util.FSUtils;
+import org.apache.hadoop.ipc.RemoteException;
+import org.apache.yetus.audience.InterfaceAudience;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import org.apache.hbase.thirdparty.org.apache.commons.collections4.MapUtils;
+
+@InterfaceAudience.Private
+public abstract class BaseRecoveredEditsOutputSink extends OutputSink {
+  private static final Logger LOG = LoggerFactory.getLogger(RecoveredEditsOutputSink.class);
+  private WALSplitter walSplitter;
+  private FileSystem walFS;
+  private Configuration conf;
+  private final ConcurrentHashMap<String, Long> regionMaximumEditLogSeqNum =
+      new ConcurrentHashMap<>();
+
+  public BaseRecoveredEditsOutputSink(WALSplitter walSplitter,
+      WALSplitter.PipelineController controller, EntryBuffers entryBuffers, int numWriters) {
+    super(controller, entryBuffers, numWriters);
+    this.walSplitter = walSplitter;
+    this.walFS = walSplitter.walFS;
+    this.conf = walSplitter.conf;
+  }
+
+  /**
+   * @return a path with a write for that path. caller should close.
+   */
+  protected RecoveredEditsWriter createWriter(byte[] region, WAL.Entry entry) throws IOException {
+    // If we already decided that this region doesn't get any output
+    // we don't need to check again.
+    if (blacklistedRegions.contains(region)) {
+      return null;
+    }
+    String tmpDirName = walSplitter.conf.get(HConstants.TEMPORARY_FS_DIRECTORY_KEY,
+        HConstants.DEFAULT_TEMPORARY_HDFS_DIRECTORY);
+    Path regionedits = getRegionSplitEditsPath(entry,
+        walSplitter.getFileBeingSplit().getPath().getName(), tmpDirName, conf);
+    if (regionedits == null) {
+      blacklistedRegions.add(region);
+      return null;
+    }
+    FileSystem walFs = FSUtils.getWALFileSystem(conf);
+    if (walFs.exists(regionedits)) {
+      LOG.warn("Found old edits file. It could be the "
+          + "result of a previous failed split attempt. Deleting " + regionedits + ", length="
+          + walFs.getFileStatus(regionedits).getLen());
+      if (!walFs.delete(regionedits, false)) {
+        LOG.warn("Failed delete of old {}", regionedits);
+      }
+    }
+    WALProvider.Writer w = walSplitter.createWriter(regionedits);
+    LOG.info("Creating recovered edits writer path={}", regionedits);
+    return new RecoveredEditsWriter(regionedits, w, entry.getKey().getSequenceId());
+  }
+
+  protected Path closeWriter(String encodedRegionName, RecoveredEditsWriter wap,
+      List<IOException> thrown) throws IOException {
+    try {
+      wap.writer.close();
+    } catch (IOException ioe) {
+      LOG.error("Could not close recovered edits at {}", wap.path, ioe);
+      thrown.add(ioe);
+      return null;
+    }
+    LOG.info("Closed recovered edits writer path={} (wrote {} edits, skipped {} edits in {} ms",
+        wap.path, wap.editsWritten, wap.editsSkipped, wap.nanosSpent / 1000 / 1000);
+    if (wap.editsWritten == 0) {
+      // just remove the empty recovered.edits file
+      if (walFS.exists(wap.path) && !walFS.delete(wap.path, false)) {
+        LOG.warn("Failed deleting empty {}", wap.path);
+        throw new IOException("Failed deleting empty  " + wap.path);
+      }
+      return null;
+    }
+
+    Path dst = getCompletedRecoveredEditsFilePath(wap.path,
+        regionMaximumEditLogSeqNum.get(encodedRegionName));
+    try {
+      if (!dst.equals(wap.path) && walFS.exists(dst)) {
+        deleteOneWithFewerEntries(wap, dst);
+      }
+      // Skip the unit tests which create a splitter that reads and
+      // writes the data without touching disk.
+      // TestHLogSplit#testThreading is an example.
+      if (walFS.exists(wap.path)) {
+        if (!walFS.rename(wap.path, dst)) {
+          throw new IOException("Failed renaming recovered edits " + wap.path + " to " + dst);
+        }
+        LOG.info("Rename recovered edits {} to {}", wap.path, dst);
+      }
+    } catch (IOException ioe) {
+      LOG.error("Could not rename recovered edits {} to {}", wap.path, dst, ioe);
+      thrown.add(ioe);
+      return null;
+    }
+    return dst;
+  }
+
+  @Override
+  public boolean keepRegionEvent(WAL.Entry entry) {
 
 Review comment:
   It is a abstract method in OutputSink. Here is an impl and only keep the compaction event. It was used in WALSplitter#splitLogFile method. 
   `// Don't send Compaction/Close/Open region events to recovered edit type sinks.`
   `        if (entry.getEdit().isMetaEdit() && !outputSink.keepRegionEvent(entry)) {`
   `          editsSkipped++;`
   `          continue;`
   `        }`
   `       entryBuffers.appendEntry(entry);`
   It need the caller to call isMetaEdit first...

----------------------------------------------------------------
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:
users@infra.apache.org


With regards,
Apache Git Services