[
https://issues.apache.org/jira/browse/CASSANDRA-8989?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14369670#comment-14369670
]
Sylvain Lebresne commented on CASSANDRA-8989:
---------------------------------------------
That's kind of a duplicate of CASSANDRA-6863: we don't properly diff range
tombstones in 2.0 so they if a read include a range tombstone, that range
tombstone will be read-repaired every time on every replica. Don't remember why
it was decided to not include it in 2.0, but if we were worried about upgrade
then, we probably want to worry even more now, and since nobody has really
complained about that for a year, maybe it's not that big a deal in practice
and maybe "upgrade to 2.1 if that's a problem for you" is a good enough answer.
> Reading from table which contains collection type using token function and
> with CL > ONE causes overwhelming writes to replicas
> -------------------------------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-8989
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8989
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Miroslaw Partyka
> Assignee: Carl Yeksigian
> Priority: Critical
> Attachments: trace.txt
>
>
> When reading from a table at the aforementioned conditions, each read from
> replica also casues write to the replica.
> Confimed in version 2.0.12 & 2.0.13, version 2.1.3 seems ok.
> To reproduce:
> {code}CREATE KEYSPACE test WITH replication = {'class':
> 'NetworkTopologyStrategy', 'DC1': 2};
> USE test;
> CREATE TABLE bug(id int PRIMARY KEY, val map<int,int>);
> INSERT INTO bug(id, val) VALUES (1, {2: 3});
> CONSISTENCY LOCAL_QUORUM
> TRACING ON
> SELECT * FROM bug WHERE token(id) <= 0;{code}
> trace contains twice:
> Appending to commitlog
> Adding to bug memtable
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
