Website: Classifieds website (customers may put advertisements, search advertisements etc)

I intend to use SOLR for searching after which return results as ID nr:s only, after which use individuals ID nr:s and query mysql, after which lastly display the outcomes with individuals ID:s.

Presently I've around 30 tables in MySQL, one for every category.

1- Do you consider I ought to get it done in a different way than above?

2- Must I only use one SOLR document, or multiple documents? Also, is document just like a SOLR index?

3- Will it be easier to Just use SOLR and skip MySQL understanding that I've a lot of posts in every table? Personally I'm far better at using MySQL than SOLR.

4- The user wants to find cars inside a specific region, how's this kind of querying carried out/completed in SOLR? Ex: q=cars&region=washington possible?

It may seem there's a lot of information on SOLR available, but there is not, and particularly not about using PHP with SOLR along with a SOLR php client... Maybe I'll write something after i have discovered all of this... Or possibly among you can write something up!

Many thanks for those help...

First, the definitions: a Solr/Lucene document is roughly something like a database row. A catalog is roughly just like a database table.

I suggest attempting to store all of the classified-related information in Solr. Querying Solr and so the database is inefficient and incredibly likely unnecessary.

Querying inside a specific region could be something similar to q=cars+region:washington presuming you've got a region area in Solr.

The Solr wiki has a lot of reliable information along with a pretty good basic tutorial. Obviously this could continually be enhanced, if you find something that is not obvious please allow the Solr team learn about it.

I can not discuss the PHP client since i have avoid using PHP.

Solr will return it's produces a syntax easily parsible using SimpleXml. You might make use of the SolPHP client library: http://wiki.apache.org/solr/SolPHP.

Solr is actually quite efficient. I would recommend putting just as much data to your Solr index as essential to retrieve my way through one hit from Solr. This might mean a smaller amount database traffic for you personally.

If you have installed the example Solr application (includes Jetty), you'll be able to develop Solr queries while using admin interface. The URI of it makes sense virtually what you would be creating in PHP.

Probably the most difficult part when starting with Solr gets the solrconfig.xml and also the schema.xml files correct. I would recommend beginning having a very fundamental config, and restart your internet application every time you give a area. Beginning served by the entire schema.xml could be confusing.

2- Must I only use one SOLR document, or multiple documents? Also, is document the just like a SOLR index?

3- Will it be easier to Just use SOLR and skip MySQL understanding that I've a lot of posts in every table? Personally I'm far better at using MySQL than SOLR.

A document is "a caseInch of solr index. Consider that you could build just one solr index per solr Core. A core functions being an independent solr Server in to the same solr insallation.

http://wiki.apache.org/solr/CoreAdmin

Years old can take shape one index merging some table contents plus some other indexes to do second level searches...

can you give more particulars regarding your architecture and data??