Yonik, to be more specific, I'm not using a custom request handler. I've
just got a Quartz job running periodically inside the Solr webapp that goes
out and pulls down updates to Solr from our database. I just expanded the
solr .war file and created a custom version with the Spring jars and my own
code added in. As for what I did to fix my open file handle issue:
Old code with the stale file handles:
public static String getLastTimestamp(String timestampField) throws
IOException
{
Map<String, String[]> args = new HashMap<String, String[]>();
args.put("q", new String[] {"*:*"}); // Query all documents
args.put("sort", new String[] {timestampField + " desc"}); // Sort
by timestamp, descending
args.put("rows", new String[] {"1"}); // Just return the top
document
SolrCore core = SolrCore.getSolrCore();
SolrQueryRequest request = new LocalSolrQueryRequest(core, args);
SolrRequestHandler handler = core.getRequestHandler(
request.getQueryType());
SolrQueryResponse response = new SolrQueryResponse();
core.execute(handler, request, response);
DocSlice docSlice = (DocSlice) response.getValues().get("response");
SolrIndexSearcher searcher = request.getSearcher();
Integer docId = docSlice.iterator().next();
Set<String> fields = new HashSet<String>();
fields.add(timestampField); // Just return the timestamp field
values
Document doc = searcher.doc(docId, fields);
String timestamp = doc.get(timestampField);
return timestamp;
}
New code:
public static String getLastTimestamp(String timestampField) throws
IOException
{
Map<String, String[]> args = new HashMap<String, String[]>();
args.put("q", new String[] {"*:*"}); // Query all documents
args.put("sort", new String[] {timestampField + " desc"}); // Sort
by timestamp, descending
args.put("rows", new String[] {"1"}); // Just return the top
document
SolrCore core = SolrCore.getSolrCore();
SolrQueryRequest request = new LocalSolrQueryRequest(core, args);
SolrRequestHandler handler = core.getRequestHandler(
request.getQueryType());
SolrQueryResponse response = new SolrQueryResponse();
core.execute(handler, request, response);
DocSlice docSlice = (DocSlice) response.getValues().get("response");
SolrIndexSearcher searcher = request.getSearcher();
Integer docId = docSlice.iterator().next();
Set<String> fields = new HashSet<String>();
fields.add(timestampField); // Just return the timestamp field
values
Document doc = searcher.doc(docId, fields);
String timestamp = doc.get(timestampField);
request.close();
searcher.close();
return timestamp;
}
Mark
On Dec 26, 2007 11:26 AM, Yonik Seeley <[EMAIL PROTECTED]> wrote:
> On Dec 26, 2007 11:16 AM, Mark Baird <[EMAIL PROTECTED]> wrote:
> > Not sure, but I believe I just found the problem. It has to do with the
> way
> > reference counts are used in the code. Once Yonik mentioned the
> reference
> > count I looked through the Solr source to see how that worked. Then I
> > looked through the custom code I've added to my Solr install and found
> an
> > instance of SolrQueryRequest that I wasn't calling close() on when done.
>
> Solr handles all the ref-counting when you use a custom request handler.
> Are you using embedded Solr?
>
> -Yonik
>
