Uploaded image for project: 'Hadoop HDFS'
  1. Hadoop HDFS
  2. HDFS-15566

NN restart fails after RollingUpgrade from 3.1.3/3.2.1 to 3.3.0

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Blocker
    • Resolution: Fixed
    • 3.3.1, 3.4.0
    • 3.3.1, 3.4.0
    • namanode
    • None

    Description

      • After rollingUpgrade NN from 3.1.3/3.2.1 to 3.3.0, if the NN is restarted, it fails while replaying edit logs.
      • HDFS-14922HDFS-14924, and HDFS-15054 introduced the modification time bits to the editLog transactions.
      • When NN is restarted and the edit logs are replayed, the NN reads the old layout version from the editLog file. When parsing the transactions, it assumes that the transactions are also from the previous layout and hence skips parsing the modification time bits.
      • This cascades into reading the wrong set of bits for other fields and leads to NN shutting down.
      2020-09-07 19:34:42,085 | DEBUG | main | Stopping client | Client.java:1361
      2020-09-07 19:34:42,087 | ERROR | main | Failed to start namenode. | NameNode.java:1751
      java.lang.IllegalArgumentException
       at com.google.common.base.Preconditions.checkArgument(Preconditions.java:72)
       at org.apache.hadoop.ipc.ClientId.toString(ClientId.java:56)
       at org.apache.hadoop.hdfs.server.namenode.FSEditLogOp.appendRpcIdsToString(FSEditLogOp.java:318)
       at org.apache.hadoop.hdfs.server.namenode.FSEditLogOp.access$700(FSEditLogOp.java:153)
       at org.apache.hadoop.hdfs.server.namenode.FSEditLogOp$DeleteSnapshotOp.toString(FSEditLogOp.java:3606)
       at java.lang.String.valueOf(String.java:2994)
       at java.lang.StringBuilder.append(StringBuilder.java:131)
       at org.apache.hadoop.hdfs.server.namenode.FSEditLogLoader.loadEditRecords(FSEditLogLoader.java:305)
       at org.apache.hadoop.hdfs.server.namenode.FSEditLogLoader.loadFSEdits(FSEditLogLoader.java:188)
       at org.apache.hadoop.hdfs.server.namenode.FSImage.loadEdits(FSImage.java:932)
       at org.apache.hadoop.hdfs.server.namenode.FSImage.loadFSImage(FSImage.java:779)
       at org.apache.hadoop.hdfs.server.namenode.FSImage.recoverTransitionRead(FSImage.java:337)
       at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.loadFSImage(FSNamesystem.java:1136)
       at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.loadFromDisk(FSNamesystem.java:742)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.loadNamesystem(NameNode.java:654)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.initialize(NameNode.java:716)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.<init>(NameNode.java:959)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.<init>(NameNode.java:932)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.createNameNode(NameNode.java:1674)
       at org.apache.hadoop.hdfs.server.namenode.NameNode.main(NameNode.java:1744)

      Attachments

        1. HDFS-15566-003.patch
          7 kB
          Brahma Reddy Battula
        2. HDFS-15566-002.patch
          5 kB
          Brahma Reddy Battula
        3. HDFS-15566-001.patch
          3 kB
          Brahma Reddy Battula

        Issue Links

          Activity

            People

              brahmareddy Brahma Reddy Battula
              brahmareddy Brahma Reddy Battula
              Votes:
              0 Vote for this issue
              Watchers:
              15 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: