I'm using the DIH to index records from a relational database. No problems,
everything works great. But now, due to the size of index (70GB w/ 25M+
docs) I need to shard and want the DIH to distribute documents evenly
between two shards. Current approach is to modify the sql query in the
config file to get only even numbered ids on one host and odd numbered ids
on the other host. Is there are more elegant way to distribute the
documents? Has anyone else come up with a better way to approach this?

Thanks,
-Jay
  • DIH: Distributing docs... Jay Hill

Reply via email to