lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alessandro Benedetti <benedetti.ale...@gmail.com>
Subject Re: Solr "search in different servers based on search keyword"
Date Mon, 13 Jul 2015 09:10:46 GMT
Hi Arijit,
let me clarify some points, ok?


2015-07-13 6:22 GMT+01:00 Arijit Saha <arijitsaha101@gmail.com>:

> Hi Solr/ lucene Experts,
>
> We are planning to build a solr/ lucene search application.
>
> As per design requirement, the files (on which search operation require to
> be done) will be lying in separate server.
>
Ok, so the datasources for you search engine, your source of information
will be files on different servers.
This is perfectly fine.
Lucene/Solr don't use the physical files you want to index for search.
You feed Solr with the Documents, which will be indexed, producing an
Inverted index and a set of inherent data structures to provide Search at
query time.
What you do really care is whether the index(es) related to your corpus of
Documents will be or not distributed across different nodes.


>
> We want to use Solr / lucene to perform search operation on files lying in
> different remote servers.
>

So, considering now the "files" to be index segments, the answer is yes.
Lucene can search between different indexes and Solr on top of it can as
well.
SolrCloud allow you to architect your Search Engine on a cluster of Solr
instances.
Each logic Collection can be partitioned in different shards ( partition of
the whole index)  and each shard can be replicated how much you want.
It is possible to implement you own routing strategy ( the way your docs go
into which shard), or use already available routing strategies.
Yo may be interested in the compositeId routing, which later applies to
your search requirement.

Take a look to those interesting docs :

https://lucidworks.com/blog/multi-level-composite-id-routing-solrcloud/
https://lucidworks.com/blog/solr-cloud-document-routing/

At indexing time you will be able to calculate the shard to send your
documents, and be able to have your documents co-located depending of a
specific key ( that can be the original server the doc is coming from)



> Do solr/ lucene support above feature of "search in different servers based
> on search keyword"
>

Now you can use at query time the same key you configured at Indexing time
and query only a subset of documents, based on their original location.

>
> I am newbie to Solr/ Lucene.
>
> Please help. Also, let know in case any additional details required.
>

Happy to help again and with better details :)

Cheers

>
>
> Much appreciated.
>
> Thanks,
> Arijit
>



-- 
--------------------------

Benedetti Alessandro
Visiting card - http://about.me/alessandro_benedetti
Blog - http://alexbenedetti.blogspot.co.uk

"Tyger, tyger burning bright
In the forests of the night,
What immortal hand or eye
Could frame thy fearful symmetry?"

William Blake - Songs of Experience -1794 England

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message