flood stage disk watermark [95%] exceeded on all indices on this node will marked read-only

elasticsearch中flood stage disk watermark [95%]  exceeded on  all indices on this node will marked read-only

在浏览器中查看索引的映射 http://localhost:9200/_all/_settings    _all或者*

显示索引为

• "meta_col_53ec17473f834426ad0cfa4fd3a2de8b": {
• "settings": {
  • "index": {
    • "number_of_shards": "5",
    • "blocks": {
      • "read_only_allow_delete": "true"
      },
    • "provided_name": "meta_col_53ec17473f834426ad0cfa4fd3a2de8b",
    • "creation_date": "1576467676781",
    • "number_of_replicas": "1",
    • "uuid": "OQDUjmvFTwiYGUye8pyxSw",
    • "version": {
      • "created": "6020199"
      }
    }
  }
• }

原因： 磁盘空间不够啦，我使用df -h查看磁盘空间，发现使用了97%，es默认达到95％就全都设置只读

解决

配置 config/elasticsearch.py linux系统中自己添加文件或者在elasticsearch.yml文件末尾添加以下内容：

# 控制洪水阶段水印。它默认为95％ cluster.routing.allocation.disk.watermark.flood_stage: 99%

或者 cluster.routing.allocation.disk.threshold_enabled: false