ES索引分片問題:
突然有一天發現es索引分片只能在集羣中的一臺服務器上。另外一臺服務器沒有索引分片。
查看es日誌發現,提示有因爲磁盤利用率處在一個high watermark 所以索引分片不能分配到該節點。
[2016-10-12 08:13:39,275][WARN ][cluster.routing.allocation.decider] [es-node-200] high disk watermark [90%] exceeded on [Qq_lSTZSTVq7EF8LHFIMfA][es-node-200][/opt/elasticsearch-2.1.1/data/evun-es/nodes/0] free: 39gb[8%], shards will be relocated away from this node [2016-10-12 08:14:09,373][INFO ][cluster.routing.allocation.decider] [es-node-200] low disk watermark [85%] exceeded on [Qq_lSTZSTVq7EF8LHFIMfA][es-node-200][/opt/elasticsearch-2.1.1/data/evun-es/nodes/0] free: 52.9gb[10.8%], replicas will not be assigned to this node [2016-10-12 08:17:09,983][INFO ][cluster.routing.allocation.decider] [es-node-200] low disk watermark [85%] exceeded on [Qq_lSTZSTVq7EF8LHFIMfA][es-node-200][/opt/elasticsearch-2.1.1/data/evun-es/nodes/0] free: 52.9gb[10.8%], replicas will not be assigned to this node
清理歷史日誌後,集羣節點索引分片開始重新路由分配,日誌如下:
[2016-10-12 08:17:40,085][INFO ][cluster.routing.allocation.decider] [es-node-200] rerouting shards: [one or more nodes has gone under the high or low watermark]
重新路由碎片:[一個或多個節點已經在高或低水位下]
由日誌提示可以看出,集羣中的任何節點處在高水位上,高水位下,都會進行索引分配的重新路由。