1> set Xmx and Xms equal. That’s not the root of our problem, but it doesn’t help either.
2> Crack the documents outside of Solr. It’s quite easy to write a SolJ program to do this, here’s an example with the reasons why it’s advisable to do this outside of Solr.: https://lucidworks.com/2012/02/14/indexing-with-solrj/ or use a service, I think the Tika project has one. Best, Erick > On Mar 18, 2019, at 8:46 AM, wclarke <wcla...@widernet.org> wrote: > > Hi, > > I am doing a recursive search on a local hard drive through DIH with Tika. > It starts slow and then it dies at Thread-14 every time with a Java Heap > error. My Xmx is at 10g, my Xms is at 512m (which might be the problem). > It takes hours to fail after it has fetched over 300,000 items and processed > over 162,000. Any ideas? > > > > -- > Sent from: http://lucene.472066.n3.nabble.com/Solr-User-f472068.html
