[Building Sakai] Who is using the Search Tool?

[Seth Theriault]

Johns, Cheryl wrote:

> We have a cluster of 5 servers and run against an Oracle db.

We also run Search (under 2.5.x) on 2 nodes with "resource" 
content stored on the file system. Our global indexes/journals 
are on a shared NFS space a la:

sharedJournalBase at org.sakaiproject.search.api.JournalSettings=/sakai/search/searchjournal

There were once issues with the search journaling and NFS, but 
I think those have been solved. In addition, up until a year 
ago, I would routinely clear the SEARCH_* tables and force a 
reindexing at each Sakai restart. That doesn't seem to be 
necessary anymore, but that could just be luck.

> I'm guessing we would have to have a separate server so as 
> not to impact performance on our servers.

I know that Michigan has run a search server in the past and 
might still do so.

You might be interested in this feature introduced in 2.6.1:

http://jira.sakaiproject.org/browse/SAK-15618

It seems to allow finer control over which servers index 
content and merge back to the central index.

Seth