Solr vs Elasticsearch


(findbestopensource) #1

I have written an article Solr vs Elasticsearch

Please post your feedback.

Regards
Aditya


(Karussell) #2

solr is not (really) distributed or at least not out of the box ...

"Both are standalone application and both are scalable and
distributed."

for in-depth comparisons can be found here:

On 14 Okt., 08:35, findbestopensource findbestopensou...@gmail.com
wrote:

I have written an article Solr vs Elasticsearchhttp://www.findbestopensource.com/article-detail/solr-vs-elasticsearch

Please post your feedback.

Regards
Aditya


(Karel Minarik) #3

Is it really an article? I see a bunch of bullet points and a link-
bait title.

Karel

On Oct 14, 8:35 am, findbestopensource findbestopensou...@gmail.com
wrote:

I have written an article Solr vs Elasticsearchhttp://www.findbestopensource.com/article-detail/solr-vs-elasticsearch

Please post your feedback.

Regards
Aditya


(Andy-2) #4

"Solr - Supports near real time search."

The current version of Solr does not support near real time search.
NRT is scheduled to be included in the next version of Solr (4.0)

On Oct 14, 2:35 am, findbestopensource findbestopensou...@gmail.com
wrote:

I have written an article Solr vs Elasticsearchhttp://www.findbestopensource.com/article-detail/solr-vs-elasticsearch

Please post your feedback.

Regards
Aditya


(Mohit Anchlia) #5

Also I think the way Solr shards is different than elasticsearch.

On Sat, Oct 15, 2011 at 1:11 AM, Andy selforganized@gmail.com wrote:

"Solr - Supports near real time search."

The current version of Solr does not support near real time search.
NRT is scheduled to be included in the next version of Solr (4.0)

On Oct 14, 2:35 am, findbestopensource findbestopensou...@gmail.com
wrote:

I have written an article Solr vs Elasticsearchhttp://www.findbestopensource.com/article-detail/solr-vs-elasticsearch

Please post your feedback.

Regards
Aditya


(Bryan Green) #6

Yes, not a very helpful article at all. Elasticseach is superior versus
SOLR. I try to get away from XML hell. JSON is a great package for objects
and it's everywhere in elasticsearch.

I feel elasticsearch has a very rich future. I'm worried though that if
kimchy were to stop developing we'd have fewer updates/features going
forward.


(Shay Banon) #7

I feel elasticsearch has a very rich future. I'm worried though that if
kimchy were to stop developing we'd have fewer updates/features going
forward.

Thats a valid concern. The way I see it, elasticsearch is now in the phase
where it starts to get noticed and used broadly. Obviously, I still play a
major role in its future, and plan to continue and doing so (not mentioning
even the time spent doing it, just my emotions associated with it...), but I
do hope for a broader developer community (user community and external
developer community (clients, integrations) have exceeded my wildest
expectations).

The stage of elasticsearch is similar in nature, now, to the state of Rails
about a year after dhh released it to the public, or more current example,
Salvatore releasing redis. I specifically mentioned those two projects as
the creators of them play a major role in the direction the projects are
heading, which I plan to do. Both are also very successful projects, which I
hope elasticsearch is on its way to become one.

I am encouraged by the increasing amount of pull requests and discussions on
IRC. Remember, elasticsearch is quite a complex project (though I hope its
not that obvious to the user ...), and combines two quite broad subjects in
both search engines (Lucene) and distributed systems. So it can be
intimidating to even start and provide pull requests for it. But, I am here
to help (with many IRC hours logged helping out with people implementing
features :wink: ), though, I expect people who want to contribute to exhibit
some form of prior effort in getting familiar with the codebase ;).

One interesting thing that I found out is that elasticsearch is being used
in quite big projects, and big / "famous" companies. Some even have
developers doing mostly elasticsearch. The reason why its not known, and why
there aren't enough public contributions to elasticsearch coming from them,
is because they consider elasticsearch as a competitive advantage. This, in
turn, will pass as elasticsearch becomes more and more popular, that
"keeping it a secret" does not make sense, but its still very encouraging
regarding the future of elasticsearch. (and a suggestion to someone who is
doing open source, don't write it in a modular fashion so it can be extended
easily ;), funnily enough, a project that is not modular, i.e. Cassandra,
results in more changes being applied to open source...).

-shay.banon


(Karussell) #8

and a suggestion to someone who is
doing open source, don't write it in a modular fashion so it can be extended
easily ;), funnily enough, a project that is not modular, i.e. Cassandra,
results in more changes being applied to open source...).

haha :slight_smile: !


(egaumer) #9

"The reason why its not known, and why there aren't enough public
contributions to elasticsearch coming from them, is because they consider
elasticsearch as a competitive advantage."

*
*
Absolutely agree.

Another interesting factor is that elasticsearch really takes search beyond
anything we've seen before. You've got document sources, versioning,
realtime GET, zero-conf network discovery, etc. With that said, ES allows
you to build higher levels of abstraction on top of search. An analogy that
always comes to mind is WordPress to MySQL. Elasticsearch embeds so easily
that you can build really interesting applications using it as the core
storage engine.

The industry terminology for this is Search Based Applications (SBAs).

This example is built completely on top of elasticsearch using backbone.js -

At any rate, this means people become less focused on the core search
(elasticsearch in this case), and more focused on the applications they're
building.

Besides, Shay's pace has been so fast that he implements features before you
realize you need it :wink:


(system) #10