[CASSANDRA-2684] IntergerType uses Thrift method that attempts to unsafely access backing array of ByteBuffer and fails

• Type: Bug
• Status: Resolved
• Priority: Low
• Resolution: Fixed
• Fix Version/s: 0.7.7, 0.8.0
• Severity: Low
• Assignee: Ed Anuff
• Reporter: Ed Anuff

Authors

Ed Anuff

• Created: 22/May/11 20:43
• Updated: 16/Apr/19 09:32
• Resolved: 23/May/11 01:34

Description

I get the following exception:

ERROR 13:27:38,153 Fatal exception in thread Thread[ReadStage:36,5,main]
java.lang.RuntimeException: java.lang.UnsupportedOperationException
	at org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:34)
	at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
	at java.lang.Thread.run(Thread.java:680)
Caused by: java.lang.UnsupportedOperationException
	at java.nio.ByteBuffer.array(ByteBuffer.java:940)
	at org.apache.thrift.TBaseHelper.byteBufferToByteArray(TBaseHelper.java:264)
	at org.apache.thrift.TBaseHelper.byteBufferToByteArray(TBaseHelper.java:251)
	at org.apache.cassandra.db.marshal.IntegerType.getString(IntegerType.java:136)
	at org.apache.cassandra.db.marshal.AbstractCompositeType.getString(AbstractCompositeType.java:131)
	at org.apache.cassandra.db.Column.getString(Column.java:228)
	at org.apache.cassandra.db.filter.SliceQueryFilter.collectReducedColumns(SliceQueryFilter.java:123)
	at org.apache.cassandra.db.filter.QueryFilter.collectCollatedColumns(QueryFilter.java:130)
	at org.apache.cassandra.db.ColumnFamilyStore.getTopLevelColumns(ColumnFamilyStore.java:1303)
	at org.apache.cassandra.db.ColumnFamilyStore.getColumnFamily(ColumnFamilyStore.java:1188)
	at org.apache.cassandra.db.ColumnFamilyStore.getColumnFamily(ColumnFamilyStore.java:1145)
	at org.apache.cassandra.db.Table.getRow(Table.java:385)
	at org.apache.cassandra.db.SliceFromReadCommand.getRow(SliceFromReadCommand.java:61)
	at org.apache.cassandra.service.StorageProxy$LocalReadRunnable.runMayThrow(StorageProxy.java:641)
	at org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:30)
	... 3 more

Tracing it down, I find that IntegerType's getString method() looks like this:

IntegerType.java

    public String getString(ByteBuffer bytes)
    {
        if (bytes == null)
            return "null";
        if (bytes.remaining() == 0)
            return "empty";

        return new java.math.BigInteger(TBaseHelper.byteBufferToByteArray(bytes)).toString(10);
    }

TBaseHelper.byteBufferToByteArray() looks like this:

TBaseHelper.java

  public static byte[] byteBufferToByteArray(ByteBuffer byteBuffer) {
    if (wrapsFullArray(byteBuffer)) {
      return byteBuffer.array();
    }
    byte[] target = new byte[byteBuffer.remaining()];
    byteBufferToByteArray(byteBuffer, target, 0);
    return target;
  }

  public static boolean wrapsFullArray(ByteBuffer byteBuffer) {
    return byteBuffer.hasArray()
      && byteBuffer.position() == 0
      && byteBuffer.arrayOffset() == 0
      && byteBuffer.remaining() == byteBuffer.capacity();
  }

  public static int byteBufferToByteArray(ByteBuffer byteBuffer, byte[] target, int offset) {
    int remaining = byteBuffer.remaining();
    System.arraycopy(byteBuffer.array(),
        byteBuffer.arrayOffset() + byteBuffer.position(),
        target,
        offset,
        remaining);
    return remaining;
  }

The second overloaded implementation of byteBufferToByteArray is calling the bytebuffer's array() method.

Suggested fixes:

1) Don't use TBaseHelper in IntegerType.getString(), use ByteBufferUtil.getArray()

2) Report problem upstream to Thrift.

3) Find a better way to deserialize BigIntegers that doesn't require an array copy.