Fwd: [DISCUSS][UPDATE] Durable Query Model Proposal

["Gang(Gary) Wang" <ga...@apache.org>]

Hi Devs.

The following ticket has been resolved, it defined the interface for
durable memory service based query engine, the user could use this low
level engine to query their durable objects in very efficient way, i.e.
using native code to filter out business durable objects from large
capacity of durable memory service. it also is flexible for the underlying
implementation of query engine since it does not assume how the query
engine works. so for example, memory service provider could simply make use
of SQLite as query engine or use customized one to avoid copying from
durable objects.

https://issues.apache.org/jira/browse/MNEMONIC-326

Best Regards.
Gary.



---------- Forwarded message ----------
Hi Devs.

Mnemonic has been evolved to be able to support more sophistical features,
so I propose to design and implement another new model for Mnemonic that is
Durable Query Model (DQM).

DQM would build on top of durable collection and memory services. DQM is
also able to take fully advantages of Durable Object Model (DOM) and
Durable Computing Model (DCM) for high performance querying and updating on
Massive Linked Business Object Sets (MLBOS).

DQM should allows to distribute and sharing all or part of MLBOS across
nodes even data centers without heavily depending on frameworks to burden
developers.

The following diagram shows the concept architecture of this new model


​
​
The applications could use 3 querying forms with hints to query their
MLBOS. the model should provide enough flexibility to accommodate best
Query Engines working on durable collections and durable memory services.

The Collection Query Engine is inspired by CQEngine
https://github.com/npgall/cqengine
Thanks Niall Gallagher for join us and help me to learn his project

There are lot we can create for DQM around durable models with various
platforms/technologies e.g. cloud, container, hpc.

The DQM would be next big feature for Mnemonic, Please review arch. any
idea, suggestion and feedback are warmly welcome, Thanks!

Very truly yours.
Gary

On Wed, Jul 26, 2017 at 12:16 PM, Johnu <jo...@gmail.com> wrote:

> Hi Gary,
>             I think, there is some error in the attachment. I am not able
> to view it.
>
> Thanks,
> Johnu
>
> On Wed, Jul 26, 2017 at 11:21 AM, Gang(Gary) Wang <ga...@apache.org>
> wrote:
>
> > Hi Devs.
> >
> > Mnemonic has been evolved to be able to support more sophistical
> features,
> > so I propose to design and implement another new model for Mnemonic that
> is
> > Durable Query Model (DQM).
> >
> > DQM would build on top of durable collection and memory services. DQM is
> > also able to take fully advantages of Durable Object Model (DOM) and
> > Durable Computing Model (DCM) for high performance querying and updating
> on
> > Massive Linked Business Object Sets (MLBOS).
> >
> > DQM should allows to distribute and sharing all or part of MLBOS across
> > nodes even data centers without heavily depending on frameworks to burden
> > developers.
> >
> > The following diagram shows the concept architecture of this new model
> >
> >
> > ​
> > The applications could use 3 querying forms with hints to query their
> > MLBOS. the model should provide enough flexibility to accommodate best
> > Query Engines working on durable collections and durable memory services.
> >
> > The Collection Query Engine is inspired by CQEngine
> > https://github.com/npgall/cqengine
> > Thanks Niall Gallagher for join us and help me to learn his project
> >
> > There are lot we can create for DQM around durable models with various
> > platforms/technologies e.g. cloud, container, hpc.
> >
> > The DQM would be next big feature for Mnemonic, Please review arch. any
> > idea, suggestion and feedback are warmly welcome, Thanks!
> >
> > Very truly yours.
> > Gary
> >
> >
> >
>