Fabien Rousseau created CASSANDRA-11477:
-------------------------------------------
Summary: MerkleTree mismatch when a cell is shadowed by a range
tombstone in different SSTables
Key: CASSANDRA-11477
URL: https://issues.apache.org/jira/browse/CASSANDRA-11477
Project: Cassandra
Issue Type: Bug
Reporter: Fabien Rousseau
Assignee: Fabien Rousseau
Below is a script which allows to reproduce the problem:
{noformat}
ccm create test -v 2.1.13 -n 2 -s
ccm node1 cqlsh
CREATE KEYSPACE test_rt WITH replication = {'class': 'SimpleStrategy',
'replication_factor': 2};
USE test_rt;
CREATE TABLE IF NOT EXISTS table1 (
c1 text,
c2 text,
c3 text,
PRIMARY KEY ((c1), c2)
);
INSERT INTO table1 (c1, c2, c3) VALUES ( 'a', 'b', 'c');
ctrl ^d
# now flush only one of the two nodes
ccm node1 flush
ccm node1 cqlsh
USE test_rt;
DELETE FROM table1 WHERE c1 = 'a' AND c2 = 'b';
ctrl ^d
ccm node1 repair test_rt table1
# now grep the log and observe that there was some inconstencies detected
between nodes (while it shouldn't have detected any)
ccm node1 showlog | grep "out of sync"
{noformat}
The wrong hash will be computed on node1, which will include the previously
deleted cell, thus resulting in a MT mismatch.
This is due to the fact that, in LazilyCompactedRow, the RT is not added into
the RT tracker (in fact, it's added only if it is GCable, but should always be
added).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
