[GEODE-8846] Transaction commit fail when uniqueID integer overflow to negative value

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.13.0, 1.13.1
• Fix Version/s: 1.14.0
• Component/s: client/server
• Labels: pull-request-available
• Assignee: Jakov Varenina
• Reporter: Jakov Varenina
• Created: 19/Jan/21 08:56
• Updated: 03/Sep/21 02:24
• Resolved: 25/Jan/21 08:11

Description

When client increments uniqId above Integer.MAX_VALUE (2147483647) then due to memory overflow the uniqId is set to negative value Integer.MIN_VALUE (-2147483648). TransactionID (uniqId) is sent to the server as a part of transaction message.

public class TXManagerImpl implements CacheTransactionManager, MembershipListener {
...
  // The unique transaction ID for this Manager
  private final AtomicInteger uniqId;
....
    TXId id = new TXId(this.distributionMgrId, this.uniqId.incrementAndGet());
....

Currently server will interpret any negative value of transactionID (uniqID) as non-transactional traffic. Please notice that getTransactionID() actually retrieves uniqID value that is set by client.

/**
 * checks to see if this thread needs to masquerade as a transactional thread. clients after
 * GFE_66 should be able to start a transaction.
 *
 * @return true if thread should masquerade as a transactional thread.
 */
protected boolean shouldMasqueradeForTx(Message clientMessage,
    ServerConnection serverConnection) {
  return serverConnection.getClientVersion().isNotOlderThan(KnownVersion.GFE_66)
      && clientMessage.getTransactionId() > TXManagerImpl.NOTX; // ---> NOTX is equal -1
}

After overflow happens all subsequent commit actions from the client are rejected with exception:

[vm0] [fatal 2021/01/14 15:28:41.967 CET <ServerConnection on port 39093 Thread 2> tid=0x52] Server connection from [identity(192.168.90.23(29826:loner):48210:6c694c01,connection=1; port=48212] : Unexpected Error on server
[vm0] org.apache.geode.InternalGemFireError
[vm0] at org.apache.geode.internal.Assert.throwError(Assert.java:91)
[vm0] at org.apache.geode.internal.Assert.assertTrue(Assert.java:55)
[vm0] at org.apache.geode.internal.cache.tier.sockets.command.CommitCommand.cmdExecute(CommitCommand.java:82)
[vm0] at org.apache.geode.internal.cache.tier.sockets.BaseCommand.execute(BaseCommand.java:183)
[vm0] at org.apache.geode.internal.cache.tier.sockets.ServerConnection.doNormalMessage(ServerConnection.java:848)
[vm0] at org.apache.geode.internal.cache.tier.sockets.OriginalServerConnection.doOneMessage(OriginalServerConnection.java:72)
[vm0] at org.apache.geode.internal.cache.tier.sockets.ServerConnection.run(ServerConnection.java:1214)
[vm0] at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
[vm0] at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
[vm0] at org.apache.geode.internal.cache.tier.sockets.AcceptorImpl.lambda$initializeServerConnectionThreadPool$3(AcceptorImpl.java:691)
[vm0] at org.apache.geode.logging.internal.executors.LoggingThreadFactory.lambda$newThread$0(LoggingThreadFactory.java:120)
[vm0] at java.lang.Thread.run(Thread.java:748)

Issue Links

• links to: GitHub Pull Request #5927