ES常用命令及注意事项

修改es配置文件：

sed -i 's/\<120s\>/300s/g' /opt/elasticsearch/config/elasticsearch.ymls/\<120s\>/300s/g 为完全匹配才会替换，但要注意唯一性vim编辑器中亦常用:%s/\<old\>/new/g

干掉es进程

ps -ef|grep elasticsearch|grep -v grep|awk '{print $2}'|xargs kill -9除了kill，可以用页面--action--关停，这种方式更好一些

关闭自动分配（重启是先关）

curl -XPUT 192.168.14.161:9200/_cluster/settings -d'{  "transient": {  "cluster.routing.allocation.enable": "none"       }}'

ES启动

/opt/elasticsearch/bin/elasticsearch -d

启用自动分配（多数节点上线后开启）

curl -XPUT 192.168.14.148:9200/_cluster/settings -d'{     "transient": {         "cluster.routing.allocation.enable": "all"     } }'

failed节点重试，刷新es（节点全部上线后使用）

curl -XPOST '192.168.14.148:9200/_cluster/reroute?retry_failed'

查看设置

curl -XGET '192.168.14.161:9200/_settings?pretty' 

curl -XPUT "192.168.14.161:9200/_cluster/settings " -d '{    "persistent" : {        "indices.recovery.max_bytes_per_sec" : "40mb"    }}'

设置分片副本数

curl -XPUT "http://192.168.14.161:9200/_settings " -d' { "number_of_replicas" : 1 } '

查看delayed_timeout

curl -s '192.168.14.161:9200/_all/_settings ' | grep delayed_timeout --color

设置delayed_timeout

curl -XPUT "192.168.14.161:9200/_all/_settings " -d '{"settings": {"index.unassigned.node_left.delayed_timeout": "5m"    }}'

查看UNASSIGNED的分片数

curl -s '192.168.14.161:9200/_cat/shards ' | grep UNASSIGNED

查看UNASSIGNED分片的详细信息

curl -XGET 192.168.14.161:9200/_cat/shards?h=index,shard,prirep,state,unassigned.reason| grep UNASSIGNED

查看node的值

curl '192.168.14.161:9200/_nodes/process?pretty

修复unassinged的分片（建议尝试retry_failed命令和重启集群未解决的情况下再用此方式修复）

curl -XPOST '192.168.14.161:9200/_cluster/reroute' -d '{        "commands" : [ {              "allocate" : {                  "index" : "skynet_public_content_v6",                  "shard" : 16,                  "node" : "lH7pzwfTQQOezWDgTtDxpg",                  "allow_primary" : true              }            }        ]    }'