health status index uuid pri rep docs.count docs.deleted store.size pri.store.size
green open sks QcmAeW2sToa3nqbPF7EVzQ 30 1 3101816690 1219212661 16.8tb 8.4tb
Hi All
I have a most large elasticsearch indices , the cluster has 20 nodes ,each one has 16G RAM and 1T SSD , now my cluster disk use 90% of disk . As you see I have almost 1/3 data deleted ,but now I can't merge it like (_forcemerge?only_expunge_deletes=false&max_num_segments=1&flush=true)
becase when it merge. it will use double disk storage , I have no such nodes to add.
Is there any way I can decrease my disk useage ?