Accidently ran another instance of elasticsearch on a node

Hi, while trying to set up an another process as master, i believe i for some time ran multiple instances of elasticsearch on three nodes.

On these nodes, its looks like this:
/var/lib/elasticsearch/elasticsearch/indices/0
/var/lib/elasticsearch/elasticsearch/indices/1

On my other nodes, that are fine it looks like:
/var/lib/elasticsearch/elasticsearch/indices/0

So, there is alot of data in the "1"-directory on three nodes, and these shards will not be ASSIGNED, my cluster stays yellow.

This mistake happend 1 week ago, since then i have restarted ES a couple of times, but it was just now that i got the problem.

How can i fix this?

At the moment im running 5 nodes, where 3 runs another instance of elasticsearch, as just masters