Purge Data from Volume

We are running some extensive Iron Man testing and I need to delete some of the data we are generating ~6 million jpg per day. However due to poor small io performance it's taking ages to delete the data through Gluster or NFS.

I noticed if I remove files locally it's not synchronized (which I understand). If I stop gluster and delete the files on both replicas locally then restart gluster will everything be ok?

Any recommendations apart from destroy the volume and re-create?

Thanks