cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Stupp (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-8929) Workload sampling
Date Fri, 06 Mar 2015 19:59:39 GMT


Robert Stupp commented on CASSANDRA-8929:

You're right - stress and wl-recording are somewhat similar.
cassandra-stress is a really cool tool to test defined workloads.
But IMO users will neglect to keep their stress profiles "up to date" - either lazily or due
to lack of time - however.

IMO WL recording/playback should be on top of all "layers" in C* - i.e. somewhere at the protocol
handling level. Maybe some kind of "network sniffer" just for CQL statements. Ideally something
that can record and playback using different protocol versions.
But hey - with some kind of "AI" it could also create stress-profiles after recording. Not
easy, but possible to some extend.

> Workload sampling
> -----------------
>                 Key: CASSANDRA-8929
>                 URL:
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Tools
>            Reporter: Jonathan Ellis
> Workload *recording* looks to be unworkable (CASSANDRA-6572).  We could build something
almost as useful by sampling the requests sent to a node and building a synthetic workload
with the same characteristics using the same (or anonymized) schema.

This message was sent by Atlassian JIRA

View raw message