compare backends

[Jeff Zemerick <jz...@apache.org>]

Hi,

I'm excited to have found Rya. I did a build and was able to insert and
execute queries using the MongoDB backend. In addition to the Accumulo
backend I see open JIRA tasks for a few other backends. Are there any
downsides to using the MongoDB backend instead of the original Accumulo
backend implementation? Is it a goal to add support for additional backends?

Thanks,
Jeff

Re: compare backends

[Jeff Zemerick <jz...@apache.org>]

Puja, thanks for the explanation! That helps a lot. I use AWS a lot and I
can certainly see a DynamoDB backend being useful.

Keith, I had found the Rya paper a few years back (probably early 2014, I
think) in a Google search. I was looking through the presentations from
Apache Big Data 2017 and I learned it is now an Apache project.

Thanks,
Jeff



On Fri, Jul 21, 2017 at 10:45 AM, Keith Turner <ke...@deenlo.com> wrote:

> On Thu, Jul 20, 2017 at 9:02 PM, Jeff Zemerick <jz...@apache.org>
> wrote:
> > Hi,
> >
> > I'm excited to have found Rya. I did a build and was able to insert and
>
> I'm curious, how did you find it?
>
> > execute queries using the MongoDB backend. In addition to the Accumulo
> > backend I see open JIRA tasks for a few other backends. Are there any
> > downsides to using the MongoDB backend instead of the original Accumulo
> > backend implementation? Is it a goal to add support for additional
> backends?
> >
> > Thanks,
> > Jeff
>

Re: compare backends

[Keith Turner <ke...@deenlo.com>]

On Thu, Jul 20, 2017 at 9:02 PM, Jeff Zemerick <jz...@apache.org> wrote:
> Hi,
>
> I'm excited to have found Rya. I did a build and was able to insert and

I'm curious, how did you find it?

> execute queries using the MongoDB backend. In addition to the Accumulo
> backend I see open JIRA tasks for a few other backends. Are there any
> downsides to using the MongoDB backend instead of the original Accumulo
> backend implementation? Is it a goal to add support for additional backends?
>
> Thanks,
> Jeff

Re: compare backends

[Puja Valiyil <pu...@gmail.com>]

Hi Jeff,
One of the main features of Rya is it provides an open rdf compliant triplestore for nosql back ends.  There is also built in security/visibility protections at the individual triple level.  

Accumulo backed Rya was intended for extremely large dataset use cases, whereas Mongo db backed Rya was meant for smaller deployments (a single node use case).  The accumulo back end is more mature than the Mongo back end.  

There are a few other back ends that exist, but they are more science experiments than hardened software at this point.  I put together a dynamo db backend for aws use cases.   I got kind of scared about how much effort it would take for me (or someone else) to maintain it though, so the pr has never been merged.  It would also be really straight forward to put together other back ends-- I can help guide/work with someone if there's an interest to do it.  I haven't gotten a good feel for what other back ends would be popular, so I've been hesitant to put in the effort.  

Sent from my iPhone

> On Jul 20, 2017, at 9:02 PM, Jeff Zemerick <jz...@apache.org> wrote:
> 
> Hi,
> 
> I'm excited to have found Rya. I did a build and was able to insert and
> execute queries using the MongoDB backend. In addition to the Accumulo
> backend I see open JIRA tasks for a few other backends. Are there any
> downsides to using the MongoDB backend instead of the original Accumulo
> backend implementation? Is it a goal to add support for additional backends?
> 
> Thanks,
> Jeff