Reputation: 354
I have only one node on one computer and the index have 5 shards without replicas. Here are some parameters describe my elasticsearch node(healthy indexes are ignored in the following list):
GET /_cat/indices?v
health status index pri rep docs.count docs.deleted store.size pri.store.size
red open datas 5 0 344999414 0 43.9gb 43.9gb
GET _cat/shards
datas 4 p STARTED 114991132 14.6gb 127.0.0.1 Eric the Red
datas 3 p STARTED 114995287 14.6gb 127.0.0.1 Eric the Red
datas 2 p STARTED 115012995 14.6gb 127.0.0.1 Eric the Red
datas 1 p UNASSIGNED
datas 0 p UNASSIGNED
shards disk.indices disk.used disk.avail disk.total disk.percent host ip node
14 65.9gb 710gb 202.8gb 912.8gb 77 127.0.0.1 127.0.0.1 Eric the Red
3 UNASSIGNED
Upvotes: 8
Views: 28487
Reputation: 354
Although deleting created shards doesn't seem to be supported, as mentioned on the comments above, reducing the number of replicas to zero for the indexes with UNASSIGNED shards might do the job, at least for single node clusters.
PUT /{my_index}/_settings
{
"index" : {
"number_of_replicas" : 0
}
}
Upvotes: 9
Reputation: 1563
You can try deleting unassigned shard following way (Not sure though if it works for data index, works for marvel indices)
1) Install elasticsearch plugin - head. Refer Elastic Search Head Plugin Installation
2) Open your elasticsearch plugin - head URL in brwoser. From here you can easily check out which are unassigned shards and other related info. This will display infor regarding that shard.
{
"state": "UNASSIGNED",
"primary": true,
"node": null,
"relocating_node": null,
"shard": 0,
"index": ".marvel-es-2016.05.18",
"version": 0,
"unassigned_info": {
"reason": "DANGLING_INDEX_IMPORTED",
"at": "2016-05-25T05:59:50.678Z"
}
}
from here you can copy index name i.e. .marvel-es-2016.05.18
.
3) Now you can run this query in sense
DELETE .marvel-es-2016.05.18
Hope this helps !
Upvotes: 0