JavaBin

[Briggs]

Hello all,

Can someone please let me know the status of the JavaBin branch? Does
it work and are there performance improvements documented? This is
something that I suspect will improve response times substantially in
which a large number of documents are returned (50,000+).

If there is stuff to do for it I maybe be able to contribute as well.

All information is appreciated!

Briggs Thompson

[Mauricio Scheffer]

Hi Briggs,

Issue is here: http://code.google.com/p/solrnet/issues/detail?id=71

I haven't touched this since December 2009 (and I think nobody else has). The codec itself is not hard to port from Java, IIRC decoding is fully implemented and could use some more testing, encoding is not implemented at all but it's not that important.

The real work is refactoring the rest of SolrNet to allow for different codecs, and change all parsers which currently work directly against XML.

However, for big results, I think lazy parsing would make a much bigger difference than javabin.

--

Mauricio

--
You received this message because you are subscribed to the Google Groups "SolrNet" group.
To post to this group, send email to sol...@googlegroups.com.
To unsubscribe from this group, send email to solrnet+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/solrnet?hl=en.

[Briggs Thompson]

Thanks for the quick response Mauricio! I appreciate all the work you have done for this community. 

I have a few use cases in which we need to return a large number of document IDs in a single request (cannot use paging for this). After a certain threshold, the response times increase substantially in proportion to the number of documents returned. Since it is just the IDs, I am thinking of returning the results in CSV format since there is less packets to transmit. I assume I would have to write a new ResponseParser for this, but that should be easy.

You bring up a good point about the lazy parsing, though it seems that would take some significant development work to get the read ahead working properly. 

Thanks again for your input,

Briggs

[Mauricio Scheffer]

Yes, as a workaround you could hook up your own ISolrQueryResultParser which would parse CSV. Since you only need the IDs, in your implementation you could just ignore all other response parsers and ISolrQueryResults properties.

--

Mauricio