集群报红问题

Question

集群第一天所有操作没事；第二天创建新的索引报红；重启后问题解决；但是第三天创建新的索引又会报红

Answer 1

rockybean 2018-06-27 14:36:48

这个问题是 shard allocation 失败的问题，那么可以使用以下接口看下失败的原因（注意是在出问题的时候看）

GET _cluster/allocation/explain

https://www.elastic.co/guide/en/elasticsearch/reference/master/cluster-allocation-explain.html

另外再关注下 es 日志里面有无有价值的信息

• 提问者 qq_无妄_3 #1

  非常感谢老师！

• rockybean 回复 提问者 qq_无妄_3 #2

  问题定位到了吗？

• 提问者 qq_无妄_3 回复 rockybean #3

  定位到了呢；老师

Answer 2

提问者 qq_无妄_3 2018-06-28 11:01:50

回复 rockybean：定位到了那个命令出来后是这个样子
GET _cluster/allocation/explain

{
  "index": "apachelog-2015.05.17",
  "shard": 2,
  "primary": true,
  "current_state": "unassigned",
  "unassigned_info": {
    "reason": "INDEX_CREATED",
    "at": "2018-06-28T02:51:05.736Z",
    "last_allocation_status": "no"
  },
  "can_allocate": "no",
  "allocate_explanation": "cannot allocate because allocation is not permitted to any of the nodes",
  "node_allocation_decisions": [
    {
      "node_id": "zMOX1zjIRqe-AxSbvWxf2A",
      "node_name": "ctgegh03",
      "transport_address": "134.96.252.88:9300",
      "node_decision": "no",
      "weight_ranking": 1,
      "deciders": [
        {
          "decider": "enable",
          "decision": "NO",
          "explanation": "no allocations are allowed due to cluster setting [cluster.routing.allocation.enable=none]"
        }
      ]
    },
    {
      "node_id": "oaperWCYRY2djW_9lYIkFA",
      "node_name": "ctgexh01",
      "transport_address": "134.96.252.90:9300",
      "node_decision": "no",
      "weight_ranking": 2,
      "deciders": [
        {
          "decider": "enable",
          "decision": "NO",
          "explanation": "no allocations are allowed due to cluster setting [cluster.routing.allocation.enable=none]"
        }
      ]
    },
    {
      "node_id": "NDwlNyXlSLO3GI25_iQx6A",
      "node_name": "ctgegh04",
      "transport_address": "134.96.252.89:9300",
      "node_decision": "no",
      "weight_ranking": 3,
      "deciders": [
        {
          "decider": "enable",
          "decision": "NO",
          "explanation": "no allocations are allowed due to cluster setting [cluster.routing.allocation.enable=none]"
        }
      ]
    }
  ]
}
日志是
[2018-06-28T10:52:42,362][DEBUG][o.e.a.a.c.a.TransportClusterAllocationExplainAction] [ctgegh04] explaining the allocation for [ClusterAllocationExplainRequest[useAnyUnassignedShard=true,includeYesDecisions?=false], found shard [[apachelog-2015.05.17][2], node[null], [P], recovery_source[new shard recovery], s[UNASSIGNED], unassigned_info[[reason=INDEX_CREATED], at[2018-06-28T02:51:05.736Z], delayed=false, allocation_status[deciders_no]]]

我开启cluster.routing.allocation.enable=none 为all就恢复了正常

• rockybean #1

  好的，这个 api 很强大，活学活用

Answer 3

提问者 qq_无妄_3 2018-06-27 14:13:02

cluster settings

Answer 4

提问者 qq_无妄_3 2018-06-27 14:07:28

GET _cluster/settings

{
  "persistent": {},
  "transient": {}
}

Answer 5

rockybean 2018-06-27 11:13:26

你的集群配置发一下

GET _cluster/settings

• 提问者 qq_无妄_3 #1

  老师；您看；啥都没啊
  GET _cluster/settings
  
  {
    "persistent": {},
    "transient": {}
  }

Answer 6

提问者 qq_无妄_3 2018-06-27 10:33:26

版本为6.1.3