2019-09-17老师，您好
ES自带热部署的远程词典的功能，我在对远程词典添加新词后发现，的确分词结果可以正确分分割出添加的词汇，但是已经在索引库中存在的文档中包含新词的没有更新。这种情况怎么解决？如果不能解决，那他岂不是就失去了远程词典的意义了吗？
望老师可以解答我的疑惑作者回复: 更新了字典以后，你需要对现有的文档执行update by query或者reindex ，请参考相关章节的视频。
4
2019-09-30老师您好，我看视频中对于节点返回集群，导致的dangling变红，给出的处理方式是直接删除该dangling索引。那这样索引上的数据不就丢了吗？或者说，有不删除索引的解决办法吗？作者回复: dangling出现的情况是，用户删除一个索引a，删除的时候恰好nodeA 关机了，或者因为一些原因，离开了集群。这个时候索引a被删除了。然后NodeA又回来了，NodeA里面带有索引a的分片，这样的场景下会导致集群变红。而索引a本来就是用户希望delete的，只是没delete干净，所以不存在你说的丢数据的情况。
1
2019-10-22老师你好，如果遇到这种问题怎么解决呢，shard是存在的且为red
{
"index": "logstash-2019.10.18",
"shard": 1,
"primary": false,
"current_state": "unassigned",
"unassigned_info": {
"reason": "MANUAL_ALLOCATION",
"at": "2019-10-21T03:56:38.094Z",
"details": "failed shard on node [OBgTcksjRU-lzQU2jb7QdQ]: failed recovery, failure RecoveryFailedException[[logstash-2019.10.18][1]: Recovery failed from {elasticsearch-logging-1}{x6cTKNFBTS-MvF7favwBTg}{6yqlOf7oTEWpV7sAxzWfGw}{10.2.33.252}{10.2.33.252:9300} into {elasticsearch-logging-0}{OBgTcksjRU-lzQU2jb7QdQ}{06Mr39VaQSawv7ViHM3JIg}{10.2.3.223}{10.2.3.223:9300}]; nested: RemoteTransportException[[elasticsearch-logging-1][10.2.33.252:9300][internal:index/shard/recovery/start_recovery]]; nested: RecoveryEngineException[Phase[2] phase2 failed]; nested: IOException[No such device or address]; ",
"last_allocation_status": "no_attempt"
},展开
2019-10-12老师，这个问题是什么导致的？能不能结合原理讲下
{
"index" : "containers-logs-2019.08.18",
"shard" : 0,
"primary" : false,
"current_state" : "unassigned",
"unassigned_info" : {
"reason" : "ALLOCATION_FAILED",
"at" : "2019-10-12T15:45:12.702Z",
"failed_allocation_attempts" : 5,
"details" : "failed shard on node [WxlYLXMWQwuEGRpVCYIPuQ]: failed recovery, failure RecoveryFailedException[[containers-logs-2019.08.18][0]: Recovery failed from {es7_01}{hd8FLM8UQkqikNTdEBoKbA}{-R1DhY8fQiu-TZ-V2f29pw}{172.18.0.5}{172.18.0.5:9300}{ml.machine_memory=6087548928, ml.max_open_jobs=20, xpack.installed=true} into {es7_02}{WxlYLXMWQwuEGRpVCYIPuQ}{t_R3pB5JQVmxh_LES8JSFA}{172.18.0.4}{172.18.0.4:9300}{ml.machine_memory=6087548928, xpack.installed=true, ml.max_open_jobs=20}]; nested: RemoteTransportException[[es7_01][172.18.0.5:9300][internal:index/shard/recovery/start_recovery]]; nested: CircuitBreakingException[[parent] Data too large, data for [<transport_request>] would be [1017681884/970.5mb], which is larger than the limit of [986932838/941.2mb], real usage: [1017681608/970.5mb], new bytes reserved: [276/276b]]; ",
"last_allocation_status" : "no_attempt"
},展开
2019-09-25很有用，给老师点赞
2019-09-22ES集群变红变黄的常见原因及定位分析解决之道——常见故障解决经验的积累，使用es提供的集群诊断API，挺棒的分享，感谢。
2019-09-13老师，如何检测集群由于网络故障或硬盘故障导致的集群不可用？此时的集群状态可能是正常的
2019-09-11"type":"circuit_breaking_exception",
"reason":"[parent]"Data too large,data for [<http_request>] would be [4528630082/4.2gb],which is larger than the limit of [4488796569/4.1gb],real usage:[4528629696/4.2gb] new bytes reserved:[386/386b]"
我改成 我问下剩下的空间都用在哪
-Xms7g
-Xms7g
indices.breaker.fielddata.limit 60%
indices.breaker.request.limit 40%
indices.breaker.total.limit 70%
indices.fielddata.cache.size 40%
我怎么解决这个问题呀展开