Re: Possible to HTTP POST (very long) queries?
On Apr 19, 2006, at 4:10 PM, Brian Lucas wrote: 2. I can no longer 'ant compile'/'dist' the newest version(s) (04/16/06 - 04/19/06) of Solr due to the error: --- [javac] /home/root/disks/src/solr-nightly/src/java/org/apache/solr/util/ AbstractSolr TestCase.java:8: package junit.framework does not exist [javac] import junit.framework.TestCase; ... No, it's not just you. The build requires the JUnit library, which is not provided. I'm guessing that whoever put that in is using an IDE that automatically includes JUnit in the classpath. For now, you can download the jar from www.junit.org and drop it into lib/. -MB
Re: Separate config and index per webapp
On May 15, 2006, at 3:26 PM, Yonik Seeley wrote: On 5/15/06, Michael Levy <[EMAIL PROTECTED]> wrote: I'd like to use Solr for a number of separate projects running on a single Tomcat instance; I thought I would have Solr running a separate webapp for each project. I understand that the default location for the solrconfig.xml and schema.xml files and for the index are under Tomcat's current-working-directory/solr/conf I am hoping I can change the default location for each webapp. Thanks! It's not yet possible, but see this thread: http://www.mail-archive.com/solr-dev@lucene.apache.org/msg00298.html It looks like JNDI offers the easiest portable way. The main SolrServlet would then need to be modified to do a JNDI lookup and base config off of that. The way I did it is I unpacked the WAR and put my config files into WEB-INF/classes/. Then I individually edit each app's solrconfig.xml and specify a unique path for the index directory. This may not be for everyone, but since I had to unpack the WAR anyway... -MB
Re: Separate config and index per webapp
Yeah, the config that I have (based on the repo version from a couple weeks ago) has set, and it works as expected. I don't know where the indexDir came from. On May 17, 2006, at 2:56 PM, Yonik Seeley wrote: Try using the latest nightly build, and changing indexDir to dataDir: -Yonik On 5/17/06, Michael Levy <[EMAIL PROTECTED]> wrote: I copied solrconfig.xml, schema.xml, synonyms.txt, protwords.txt and stopwords.txt into the WEB-INF/classes directory and the webapp is reading those files. I was thinking that if I uncommented the indexDir element in solrconfig.xml, it would then use that directory in which to place the index The commented-out default value is: index I tried changing the value to: indexXXX However this is being ignored; the index directory is still being created under Tomcat's current-working-directory/solr/conf/index (In my case I'm testing under Windows and it's C:\Tomcat 5.5\solr\data \index) Any ideas? Thanks. Mike Baranczak wrote: > On May 15, 2006, at 3:26 PM, Yonik Seeley wrote: > >> On 5/15/06, Michael Levy <[EMAIL PROTECTED]> wrote: >>> I'd like to use Solr for a number of separate projects running on a >>> single Tomcat instance; I thought I would have Solr running a separate >>> webapp for each project. >>> >>> I understand that the default location for the solrconfig.xml and >>> schema.xml files and for the index are under Tomcat's >>> current-working-directory/solr/conf >>> >>> I am hoping I can change the default location for each webapp. Thanks! >> >> It's not yet possible, but see this thread: >> http://www.mail-archive.com/solr-dev@lucene.apache.org/ msg00298.html >> >> It looks like JNDI offers the easiest portable way. >> The main SolrServlet would then need to be modified to do a JNDI >> lookup and base config off of that. > > > The way I did it is I unpacked the WAR and put my config files into > WEB-INF/classes/. Then I individually edit each app's solrconfig.xml > and specify a unique path for the index directory. This may not be for > everyone, but since I had to unpack the WAR anyway... > > -MB
Re: Java or PHP code to trigger solr?
Here's a Java client for Solr: http://issues.apache.org/jira/browse/SOLR-20 Erik Hatcher was recently talking about a Ruby-on-Rails interface for Solr, but I don't know what the status is on that. As far as I know, nobody has yet written interfaces for PHP or CF, but it shouldn't be that hard; it's just a question of creating an XML document and sending it over HTTP. -MB On Jun 28, 2006, at 3:54 PM, Tim Archambault wrote: I'm curious as well as I am creating a CFMX interface for Solr. If you wouldn't mind sharing info, please do. On 6/28/06, UpAndGone <[EMAIL PROTECTED]> wrote: Hello, I was just planning to write a web service like front-end to Lucene when I heard of this project. Are there any examples on integrating Solr yet? I see the wiki section, but it is empty. How can I trigger solr, say from my Java app without using curl? Thanks Christian
Re: client code for searching?
Check the list archives - there was a thread just a couple of weeks ago about all the various existing Solr clients. -MB On Jul 14, 2006, at 2:21 PM, WHIRLYCOTT wrote: Does anybody have some client code for performing searches against a Solr installation? I've seen the DocumentManagerClient for adding/dropping/etc docs from the index, but I don't see any client code in svn anywhere. If there's nothing that I can pick up and start using, I'm happy to put a standalone Solr client together and contribute that. Any pointers for either of these? phil. -- Whirlycott Philip Jacob [EMAIL PROTECTED] http://www.whirlycott.com/phil/
Re: Mac OSX - error reading /usr/local/lib/libsvnjavahl-1.0.0.0.dylib
A .dylib file isn't a zip or a jar at all, it's a native OS X shared library. I have absolutely NO idea why Solr is trying to open it. Are you deploying just the stock version, or did you add some of your own code to it? -MB On Aug 15, 2006, at 2:06 AM, Ross McDonald wrote: Thanks for the quick response guys, I am using jdk 1.5, and am ensuring use of this jdk by typing the full path. As regards doing an 'unzip -l' on the file it indeed generates an error.. /usr/local/lib rossputin$ unzip -l libsvnjavahl-1.0.0.0.dylib Archive: libsvnjavahl-1.0.0.0.dylib End-of-central-directory signature not found. Either this file is not a zipfile, or it constitutes one disk of a multi-part archive. In the latter case the central directory and zipfile comment will be found on the last disk(s) of this archive. note: libsvnjavahl-1.0.0.0.dylib may be a plain executable, not an archive unzip: cannot find zipfile directory in one of libsvnjavahl-1.0.0.0.dylib or libsvnjavahl-1.0.0.0.dylib.zip, and cannot find libsvnjavahl-1.0.0.0.dylib.ZIP, period. oh dear.. maybe this is corrupt? A 'jar tvf' generates no output, once again indicating that the file is not a valid jar, regards, Ross.
redundant Solr servers
I need to set up a cluster with two or more boxes running Solr, each using a clone of the same index (for load balancing). Does anybody have experience with such a scheme? I have a few ideas of my own, but I don't want to reinvent the wheel. -MB
Re: redundant Solr servers
On Sep 15, 2006, at 5:40 PM, Mike Klaas wrote: On 9/15/06, Mike Baranczak <[EMAIL PROTECTED]> wrote: I need to set up a cluster with two or more boxes running Solr, each using a clone of the same index (for load balancing). Does anybody have experience with such a scheme? I have a few ideas of my own, but I don't want to reinvent the wheel. Solr has a replication scheme built-in: http://wiki.apache.org/solr/CollectionDistribution -Mike Wow, that was easy. This looks like exactly what I need. I had somehow completely missed that. -MB
Limit on number of schema fields?
Is there any significant penalty for having a large number of fields in a Solr schema (like between 50 and 100)? We have a site with several different types of searchable content, and each of those types will require several different fields (most of which are not shared). I figured that it'd be easier to have everything in one index, but I just want to make sure this won't cause any problems. -MB
