[
https://issues.apache.org/jira/browse/CASSANDRA-9815?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14629954#comment-14629954
]
Shawn Kumar commented on CASSANDRA-9815:
----------------------------------------
Spoke with Aleksey regarding this and it looks like this is actually the
correct behaviour and a consequence of CASSANDRA-9299. That being said we
agreed the "Read 0 live and 0 tombstone cells" message should be modified for
clarity.
> Tracing doesn't seem to display tombstones read accurately
> ----------------------------------------------------------
>
> Key: CASSANDRA-9815
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9815
> Project: Cassandra
> Issue Type: Bug
> Components: Tools
> Reporter: Shawn Kumar
> Priority: Minor
> Labels: tracing
> Fix For: 2.2.x
>
> Attachments: tracescreen.png
>
>
> It seems that where tracing
> [once|http://stackoverflow.com/questions/27063508/how-to-get-tombstone-count-for-a-cql-query]
> tracked how many tombstones were read in a query, it no longer happens.
> Can reproduce with the following:
> 1. Create a simple key, val table.
> 2. Insert a couple rows
> 3. Flush
> 4. Delete a row, add an additional couple rows
> 5. Flush
> 6. Try to query deleted row or select *.
> In the trace it never mentions reading a tombstoned cell, no matter what.
> Instead you get a line like the following: "Read 0 live and 0 tombstone cells
> [SharedPool-Worker-3]". Attached is a screenshot of the trace.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
