accumulo-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marc Parisi <>
Subject Re: Cosmos
Date Sat, 08 Feb 2014 21:39:52 GMT

    You should be able to implement jdbc calls as the tests do. I've toyed
with the idea of improving some of the SQL functionality to support more
flexibility on where operations are performed and as of a few months ago
began developing a better framework for using real time schema statistics
to optimize the query stack. If you see utility in any of this, I can push
what I had ( which was working at the time ), but it's very experimental,
and I wanted to replace my implementation of HyperLogLogPlus with
Stream-Lib (, since the last thing
the world needs is another custom implementation.

     FYI: It's developed by a fella
named Julian Hyde. He's like the beastmaster of data management.

On Sat, Feb 8, 2014 at 4:19 PM, Josh Elser <> wrote:

> Sadly, at the moment I don't think anyone is working on it. I'd like to
> make the time to work on it some more, but I haven't been able to make that
> happen. Last I remember, there were functioning XML data load and then
> basic select and filter SQL functionality working with a simple shell.
> Cosmos is just a general indexing framework that provides some basic SQL
> functionality as a query layer. To accomplish query tasks, Accumulo
> iterators are used. For SQL support, Optiq is used which provides parser
> and optimizer functionality.
> It's a neat little project; I hope to get back to it.
> On 2/8/14, 3:58 PM, Arshak Navruzyan wrote:
>> Josh or anyone,
>> Is Cosmos an active project ?
>> main/java/cosmos/sql
>> Can you please provide some background on the capabilities?  Does it
>> index data itself or is it more of a wrapper around native accumulo
>> iterators?
>> Thanks
>> Arshak

View raw message