# Why shard unassigned after cluster restart completely?
**URL:** https://discuss.elastic.co/t/why-shard-unassigned-after-cluster-restart-completely/230460
**Category:** Elasticsearch
**Created:** [April 30, 2020, 3:30am UTC](https://discuss.elastic.co/t/why-shard-unassigned-after-cluster-restart-completely/230460 "2020-04-30T03:30:56Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@wfeng888](https://discuss.elastic.co/u/wfeng888)
#### Post date: [April 30, 2020, 3:30am UTC](https://discuss.elastic.co/t/why-shard-unassigned-after-cluster-restart-completely/230460/1 "2020-04-30T03:30:57Z")
I have a cluster with 3 nodes, all have node.master: true and node.data: true. Almost every index loss one or two shards either primary or replica after i restart three nodes 1 by one. All indices have settings of 5 primary and 1 replica. There had no errors in log files. In the end, i reroute these shard by allocate\_empty\_primary . I'm looking for the reason for this by read a lot of articles , but none of them matched my situation.
es version 5.4.3.
The explain API shows:
{
"index": "dcvs\_nonmotorvehicle",
"shard": 3,
"primary": true,
"current\_state": "unassigned",
"unassigned\_info": {
"reason": "CLUSTER\_RECOVERED",
"at": "2020-04-10T03:40:41.127Z",
"last\_allocation\_status": "no\_valid\_shard\_copy"
},
"can\_allocate": "no\_valid\_shard\_copy",
"allocate\_explanation": "cannot allocate because a previous copy of the primary shard existed but can no longer be found on the nodes in the cluster",
"node\_allocation\_decisions": [
{
"node\_id": "TklXzLKySf-czdu8zZ5hyQ",
"node\_name": "host2",
"transport\_address": "host2:9300",
"node\_attributes": {
"cname": "202",
"rack\_id": "rack\_two"
},
"node\_decision": "no",
"store": {
"found": false
}
},
{
"node\_id": "WDaA85bmQhKZxvgq4ve0Kw",
"node\_name": "host3",
"transport\_address": "host3:9300",
"node\_attributes": {
"cname": "210",
"rack\_id": "rack\_three"
},
"node\_decision": "no",
"store": {
"found": false
}
},
{
"node\_id": "dAXXDGDyQaGcex6VcZP0eg",
"node\_name": "host1",
"transport\_address": "host1:9300",
"node\_attributes": {
"cname": "201",
"rack\_id": "rack\_one"
},
"node\_decision": "no",
"store": {
"found": false
}
}
]
}
---
### Author:  [@system](https://discuss.elastic.co/u/system)
#### Post date: [May 28, 2020, 3:42am UTC](https://discuss.elastic.co/t/why-shard-unassigned-after-cluster-restart-completely/230460/2 "2020-05-28T03:42:26Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.