本文为博主九师兄(QQ:541711153 欢迎来探讨技术)原创文章,未经允许博主不允许转载。
文章目录
- 1.概述
- 2. 常用命令
- 5.未分配的原因
-
- 5.1 INDEX_CREATED
- 5.2 CLUSTER_RECOVERED
- 5.3 INDEX_REOPENED
- 5.4 DANGLING_INDEX_IMPORTED
- 5.5 NEW_INDEX_RESTORED
- 5.6 EXISTING_INDEX_RESTORED
- 5.7 REPLICA_ADDED
- 5.8 ALLOCATION_FAILED
- 5.9 NODE_LEFT
- 5.10 REINITIALIZED
- 5.11 REROUTE_CANCELLED
- 5.12 REALLOCATED_REPLICA
- 5.13 PRIMARY_FAILED
- 5.14 FORCED_EMPTY_PRIMARY
- 5.15 MANUAL_ALLOCATION
- 6.Elasticsearch中解析未分配的各个原因解析
-
- 6.1、故意分配碎片分配
- 6.2、分片太多,节点不够
- 6.3、加入一个新的节点,需要重新启用分片分配
- 6.4、集群中不在存在分片数据
- 6.5、磁盘水印低(Low disk watermark)
- 6.6 主分片数据陈旧
- 6.7 主分片数据丢失
- 6.8 yellow状态集群完整,但副分片离线
-
- 6.8.1 initializing_shards > 0
- 6.8.2 initializing_shards = 0
-
- 6.8.2.1 there are too many shards on the node for attribute
- 6.8.2.2 too many shards are concurrently rebalancing
- 6.8.2.3 the node is above the high watermark even though this shard has never been allocated and has less than required [%s] free on node, free: [%s]
- 6.8.2.4 no allocations are allowed due to
- 6.8.2.5 shard has already failed allocating [10] times vs. [" + maxRetry + "] retries allowed + unassignedInfo.toString() + " - manually call [/_cluster/reroute?retry_failed=true] to retry
- 6.8.2.6 target node version [2.0] is older than the source node version [5.1.1]
- 6.8.2.7 Priority indices not started
- 6.8.2.8 the shard cannot be allocated on the same node id [%s] on which it already exists
- 6.8.2.9 too many shards for this node [%d], cluster-level limit per node: [%d]
- 6.8.2.10 too many incoming shards are currently recovering [%d], limit: [%d]
- 6.8.3 delayed_unassigned_shards > 0
- 6.9 查看线程池的入库和查询请求
- 6.10 API响应慢且FullGC
- 7.手动移动
- 10.扩展阅读
- 11 参考
![在这里插入图片描述](https://i-blog.csdnimg.cn/blog_migrate/6ccbec930d0504bcac6ae1519f1cc7cf.png#pic_center)
1.概述
分片认识参考:【Elasticsearch】elasticsearch 分片
官网:https://www.elastic.co/guide/en/elasticsearch/reference/current/cluster-allocation-explain.html
http://doc.codingdict.com/elasticsearch/225/
1.1 检查集群的健康状况
[root@ngsocdev14 ~