[
https://issues.apache.org/jira/browse/CASSANDRA-20397?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Caleb Rackliffe updated CASSANDRA-20397:
----------------------------------------
Bug Category: Parent values: Correctness(12982)Level 1 values:
Unrecoverable Corruption / Loss(13161)
Complexity: Normal
Discovered By: Adhoc Test
Fix Version/s: 4.0.x
4.1.x
5.0.x
5.x
Severity: Normal
Status: Open (was: Triage Needed)
> FQL does not handle explicit null/tombstones in prepared statements
> -------------------------------------------------------------------
>
> Key: CASSANDRA-20397
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20397
> Project: Apache Cassandra
> Issue Type: Bug
> Components: Tool/fql
> Reporter: Caleb Rackliffe
> Assignee: Caleb Rackliffe
> Priority: Normal
> Fix For: 4.0.x, 4.1.x, 5.0.x, 5.x
>
>
> FQL will throw an NPE if it attempts to record a successful batch operation
> containing a null bind value, which we obviously interpret as a tombstone.
> This means it will be missing from any subsequent dump and completely ignored
> by replay.
> {noformat}
> ERROR [node1_Binary Log thread:1] 2025-03-04 14:23:15,372 BinLog.java:271 -
> Unexpected exception in binary log thread
> java.lang.NullPointerException: null
> at net.openhft.chronicle.bytes.BytesStore.wrap(BytesStore.java:131)
> at
> org.apache.cassandra.fql.FullQueryLogger$Batch.writeMarshallablePayload(FullQueryLogger.java:459)
> at
> org.apache.cassandra.utils.binlog.BinLog$ReleaseableWriteMarshallable.writeMarshallable(BinLog.java:364)
> at
> net.openhft.chronicle.wire.MarshallableOut.writeDocument(MarshallableOut.java:152)
> at
> org.apache.cassandra.utils.binlog.BinLog.processTasks(BinLog.java:250)
> at org.apache.cassandra.utils.binlog.BinLog.run(BinLog.java:267)
> {noformat}
> The following quick test should hit it...
> {noformat}
> @Test
> public void test() throws InterruptedException
> {
> CLUSTER.get(1).nodetool("enablefullquerylog", "--path",
> temporaryFolder.getRoot().getAbsolutePath());
> com.datastax.driver.core.Cluster.Builder builder1 =
>
> com.datastax.driver.core.Cluster.builder().addContactPoint("127.0.0.1");
> try (com.datastax.driver.core.Cluster cluster1 = builder1.build();
> Session session1 = cluster1.connect())
> {
> BatchStatement batch = new
> BatchStatement(BatchStatement.Type.UNLOGGED);
> PreparedStatement preparedWrite = session1.prepare("INSERT INTO "
> + KS_TBL + " (k, c, s) VALUES ( ?, ?, ?)");
> batch.add(preparedWrite.bind(0, 0, null));
> assertTrue(session1.execute(batch).wasApplied());
> }
> CLUSTER.get(1).nodetool("disablefullquerylog");
> ToolRunner.ToolResult negativeRunner =
> ToolRunner.invokeClass("org.apache.cassandra.fqltool.FullQueryLogTool",
> "dump", temporaryFolder.getRoot().getAbsolutePath());
> System.err.println(negativeRunner.getStdout());
> }
> {noformat}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
