[CASSANDRA-7953] RangeTombstones not merging during compaction

• Type: Bug
• Status: Resolved
• Priority: Urgent
• Resolution: Fixed
• Fix Version/s: 2.1.12, 2.2.4
• Labels: compaction deletes tombstone

Environment

Cassandra 2.1

• Severity: Critical
• Assignee: Branimir Lambov
• Reporter: Marcus Olsson

Authors

Branimir Lambov

Reviewers

Marcus Eriksson

• Created: 17/Sep/14 13:30
• Updated: 16/Apr/19 09:31
• Resolved: 01/Dec/15 15:17

Description

When performing a compaction on two sstables that contain the same RangeTombstone with different timestamps, the tombstones are not merged in the new sstable.

This has been tested using cassandra 2.1 with the following table:

CREATE TABLE test(
  key text,
  column text,
  data text,
  PRIMARY KEY(key, column)
);

And then doing the following:

INSERT INTO test (key, column, data) VALUES ("1", "1", "1"); // If the sstable only contains tombstones during compaction it seems that the sstable either gets removed or isn't created (but that could probably be a separate JIRA issue).
INSERT INTO test (key, column, data) VALUES ("1", "2", "2"); // The inserts are not actually needed, since the deletes create tombstones either way.
DELETE FROM test WHERE key="1" AND column="2";

nodetool flush

INSERT INTO test (key, column, data) VALUES ("1", "2", "2");
DELETE FROM test WHERE key="1" AND column="2";

nodetool flush
nodetool compact

When checking with the SSTableExport tool two tombstones exists in the compacted sstable. This can be repeated, resulting in more and more tombstones.

Issue Links

• is duplicated by: CASSANDRA-10505 Duplicate range tombstones
• is related to: CASSANDRA-10791 RangeTombstones can be written after END_OF_ROW markers when streaming
• links to: patch for 2.0