Thanks for the heads up. I just tested this and you are right. I am making a call to "addBeans" and it succeeds without any issue even when the server is down. That sucks.
A big part of this process is reliant on knowing exactly what has made it into the index and what has not, so this a difficult problem to solve when you can't catch exceptions. I was thinking I could execute a ping request first to determine if the Solr server is still operational, but that doesn't help if the updateRequestHandler fails. On Wed, Oct 10, 2012 at 1:48 PM, Shawn Heisey <s...@elyograg.org> wrote: > On 10/9/2012 3:02 PM, Briggs Thompson wrote: > >> *Otis* - jstack is a great suggestion, thanks! The problem didn't happen >> >> this morning but next time it does I will certainly get the dump to see >> exactly where the app is swimming around. I haven't used >> StreamingUpdateSolrServer >> but I will see if that makes a difference. Are there any major drawbacks >> of >> going this route? >> > > One caveat -- when using the Streaming/Concurrent object, your application > will not be notified when there is a problem indexing. I've been told there > is a way to override a method in the object to allow trapping errors, but I > have not seen sample code and haven't figured out how to do it. I've filed > an issue and a patch to fix this. It's received some comments, but so far > nobody has decided to commit it. > > https://issues.apache.org/**jira/browse/SOLR-3284<https://issues.apache.org/jira/browse/SOLR-3284> > > Thanks, > Shawn > >
