On 12/28/12 3:54 PM, Erik Hatcher wrote:
Not so perverse, really. Quite by design that Solr depends on SolrJ.
SolrJ _is_ Solr. EmbeddedSolrServer for example. The distributed
and replication aspects also rely explicitly on SolrJ for
communication and more.
I understand Karl's situation here, though I think the only thing I
can say is use the SolrJ appropriate for your Solr version, which
means Manifold's hook may require a separate output connector, one
for each version of Solr. I think, for example, that ManifoldCF
would benefit on indexing performance and leveraging ZK discovery in
a SolrCloud environment using the CloudSolrServer API.
I would imagine, though, that simply using SolrJ in XML-only mode
would work for all the essentials for all of those Solr versions you
mentioned using the Solr 4 version of SolrJ (I've not tried this
myself though).
It's quite difficult to manage different versions (possibly conflicting)
of the same API in a single JVM. It can be done, but at a cost of great
complexity and other tradeoffs.
I think it would be very useful if it was possible to configure SolrJ to
use a safe back-compat version of the protocol that is guaranteed to
work properly with a range of older Solr versions, even if it meant some
loss of functionality or performance.
--
Best regards,
Andrzej Bialecki
http://www.sigram.com, blog http://www.sigram.com/blog
___.,___,___,___,_._. __________________<><____________________
[___||.__|__/|__||\/|: Information Retrieval, System Integration
___|||__||..\|..||..|: Contact: info at sigram dot com
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
