After fidgeting with the split shards API for a few hours, I stumbled across the line in solr documentation which states that `*split shard api can only be used for SolrCloud collections created with numShards parameter, meaning collections which rely on Solr’s hash-based routing mechanism.*`
Nevertheless, I found something interesting while trying this in implicit routing [route parameter is a field in solr document]. Both splitting and searching is happening as expected i.e. the shard I'm trying to split becomes inactive meanwhile creating 2 subshards which are being used for searching even though the route parameter contains the parent shard name. Issue comes while indexing [using solrJ 8.10, CloudHttp2SolrClient's addBeans method to index docs], as it is indexing the data in the primary shard itself and not in the sub shard. *Why is routing working fine while searching and not while indexing? Am I missing something here? Is there a way to use this in implicit routing as well?* It will be a very tedious and long process if I had to recreate a new solr collection with new shards [removing the old shard and dividing it into new shards], will have to re-index and take the new index to production. Please suggest an optimal way.
