On 2/13/2014 9:20 AM, Murugan, Muniraja (CTR) Offshore wrote:
I am getting "ORA-04030: out of process memory when trying to allocate 4032 bytes" error when I try to index xmltype data from Oracle DB. I have registered in Solr support the ID is "SOLR-5723"
The issue tracker is for bugs and feature requests. This mailing list and other discussion forums are for support. This is probably not a bug in Solr, so we'll handle it here. I will close the issue, but if it turns out that it actually is a bug, we can re-open it.
By default, JDBC tries to buffer the entire result set in RAM. Unless your Java heap size is big enough to handle everything Solr needs as well as an entire SQL result, it's going to fail.
You need the batchSize parameter on your dataSource in the dataimport config. Try a value of -1 first, if that doesn't work, try something like 500. I know for sure that a value of -1 works with a MySQL database, but I have never used Oracle, so I don't know if it will work.
You might need to get in touch with Oracle to get a different version of their JDBC driver. A bug in the driver is much more likely than one in DIH.
http://wiki.apache.org/solr/DataImportHandler#Configuring_JdbcDataSource Thanks, Shawn
