cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Corless <>
Subject Re: Modeling Time Series data
Date Sat, 12 Jan 2019 02:02:38 GMT
Hello Akash!

For Time Series, I'd make two recommendations:

   - Check out at KairosDB <>. Works on
   Cassandra as well as Scylla; they've been around for a while, and Brian
   Hawkins has spoken a great deal on KairosDB, both at Scylla Summit 2017
   and 2018
   - Also check out Newts <>. Also usable
   for both Cassandra and Scylla. This comes from the folks at OpenNMS, and
   Jesse White had a great talk at Scylla Summit 2018
   on the topic.

While you *can* do basic time series natively in Scylla or Cassandra
without such add-ons, look at these projects to do it with more emphasis on
the 'Time Series' part.

Would love to hear your thoughts, as well as the thoughts of others, if
they've done any analysis, or actually tested or gone to production with
either of these two projects.

Your friendly neighborhood,


On Fri, Jan 11, 2019 at 2:45 PM Akash Gangil <> wrote:

> Hi,
> I have a data model where the partition key for a lot of tables is based
> on time
> (year, month, day, hour)
> Would this create a hotspot in my cluster, given all the writes/reads
> would go to the same node for a given hour? Or does the cassandra storage
> engine also takes into account the table info like table name, when
> distributing the data?
> If the above model would be a problem, what's the suggested way to solve
> this? Add tablename to partition key?
> --
> Akash

Peter Corless
Technical Marketing Manager

View raw message