[
https://issues.apache.org/jira/browse/CASSANDRA-13983?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16227645#comment-16227645
]
Ariel Weisberg commented on CASSANDRA-13983:
--------------------------------------------
The goals are a bit different. Using this as the basis for load testing is not
the primary goal although eventually it will be the basis for that as well. The
primary goal is being able to test correctness (defined as do these two
versions return the same result) using your actual data and queries.
The replay tool I created previously didn't have access to the values from the
queries which is what lead to this because I found it was very difficult to
synthesize representative queries. Once I have a log of the entire query I can
adapt the a replay tool to replay the results and compare them which wasn't a
goal in CASSANDRA-6572. It's also something that isn't served by using stress
at all.
Stress can't synthesize the corner cases of data types, data values, data
sizes, and actual data generated by a previous version of Cassandra on disk.
> Support a means of logging all queries as they were invoked
> -----------------------------------------------------------
>
> Key: CASSANDRA-13983
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13983
> Project: Cassandra
> Issue Type: New Feature
> Components: CQL, Observability, Testing, Tools
> Reporter: Ariel Weisberg
> Assignee: Ariel Weisberg
> Fix For: 4.0
>
>
> For correctness testing it's useful to be able to capture production traffic
> so that it can be replayed against both the old and new versions of Cassandra
> while comparing the results.
> Implementing this functionality once inside the database is high performance
> and presents less operational complexity.
> In [this patch|https://github.com/apache/cassandra/pull/169] there is an
> implementation of a full query log that logs uses chronicle-queue (apache
> licensed, the maven artifacts are labeled incorrectly in some cases,
> dependencies are also apache licensed) to implement a rotating log of queries.
> * Single thread asynchronously writes log entries to disk to reduce impact on
> query latency
> * Heap memory usage bounded by a weighted queue with configurable maximum
> weight sitting in front of logging thread
> * If the weighted queue is full producers can be blocked or samples can be
> dropped
> * Disk utilization is bounded by deleting old log segments once a
> configurable size is reached
> * The on disk serialization uses a flexible schema binary format
> (chronicle-wire) making it easy to skip unrecognized fields, add new ones,
> and omit old ones.
> * Can be enabled and configured via JMX, disabled, and reset (delete on disk
> data), logging path is configurable via both JMX and YAML
> * Introduce new {{fqltool}} in /bin that currently implements {{Dump}} which
> can dump in a human readable format full query logs as well as follow active
> full query logs
> Follow up work:
> * Introduce new {{fqltool}} command Replay which can replay N full query logs
> to two different clusters and compare the result and check for
> inconsistencies. <- Actively working on getting this done
> * Log not just queries but their results to facilitate a comparison between
> the original query result and the replayed result. <- Really just don't have
> specific use case at the moment
> * "Consistent" query logging allowing replay to fully replicate the original
> order of execution and completion even in the face of races (including CAS).
> <- This is more speculative
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
