Hi!

Because we develop our project with ONE database, we also want to have ONE index-folder on our server, for all of aour developers.

Now each developer has its own index-files which is configured in "hibernate.cfg" as:
<property name="hibernate.search.default.directory_provider"
value="filesystem"/>

<property name="hibernate.search.default.indexBase"
value="path-to-lucene-index"/>

So how do we have to configure our project, so that every developer has access to ONE central index?
Do we have to change the vlue of indexBase into something like "http://projextname/lucene/indexes/" or how does it work?

Thnaks in advance!!!

Hi,
from the example "http://projextname/lucene/indexes/" I understand you want to share the index across multiple machines connected in network?



This confuses me, you mean a single index, on the server right? How does it related to developers, are they all connecting to a single database and running Hibernate Search locally?
Our expectation is that you run Hibernate Search on your server, not locally on the computer of your developers. That is normally only done for development, but then in such a case people don't share the same database.

If you are looking for a network-based share index, you might want to use the Infinispan Directory: http://docs.jboss.org/hibernate/search/5.0/reference/en-US/html_single/#infinispan-directories

Yes, in development each developer has access to ONE shared database, so that every developer has the same data available.
So i thought it has to be possible to access the index for that database by an url...
We now solved the problem by sharing the index-data by shared network-device.

When the application is finished we have another database and connect it with the index which is located on the same server where the database is located.
I think it is a bit complicated, but it works.

Ok, thanks that's an interesting perspective.
Lucene is usually used strictly locally, so I had difficulty to understand your use case. Your solution seems fine for development, but keep in mind that for "production use", Lucene's index management expects an accurate implementation of the locking API on files, which is usually relaxed on networked file systems. So I would only suggest this setup as a development facility, and be aware that you might have IOExceptions corrupting the index; I assume that's not a big problem.