cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-6572) Workload recording / playback
Date Wed, 01 Oct 2014 02:37:35 GMT


Jonathan Ellis commented on CASSANDRA-6572:

bq. We also might want to create a downgradesstables tool, to make trying out new versions
even easier (by making reverting back easier).

I'm even more excited about using this to get us realistic workloads to benchmark, than I
am about helping people upgrade with confidence.  So, downgradesstables is a nice idea but
I don't think it obsoletes this.

bq. If we are using this to replicate workloads for replaying a workload, then the new stress
tool can be better configured to do that. 

Okay, so maybe we shift our goals a bit.  Assuming we can tune newstress to do 50% inserts
to table X, 10% reads from table Y, 40% updates to table Z, doing that by hand is a pain.
 What if we make this tool gather statistics like that from a live cluster, then write out
a stress schema + config to emulate it?

> Workload recording / playback
> -----------------------------
>                 Key: CASSANDRA-6572
>                 URL:
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core, Tools
>            Reporter: Jonathan Ellis
>            Assignee: Carl Yeksigian
>             Fix For: 2.1.1
>         Attachments: 6572-trunk.diff
> "Write sample mode" gets us part way to testing new versions against a real world workload,
but we need an easy way to test the query side as well.

This message was sent by Atlassian JIRA

View raw message