ElasticSearch 5.6 upgrade to 6.5 caused problems

The company has been using ElasticSearch service Amazon AWS, most recently as AWS provide an upgrade from 5.6 to 6.5, you try to upgrade a bit on Sundays. The result was a bunch of questions after the upgrade, while support open ticket while their own research how to repair the AWS. Fortunately, fixes some final, Here are some questions and experiences

 

After 1. ES 6.0 is no longer supported string type:

type here refers to the index and the indice inside the property type, before I used this format is the following

 "request" : {

            "type" : "string",

            "fields" : {

              "raw" : {

                "index" : "not_analyzed",

                "type" : "string"

              }

            }

          }

Phenomenon after the upgrade is unable to create the log indice day on the ES cluster. Later found to be from the beginning because ES 6.0 is no longer supported string. So with Dev Tools upgrade template to look like

 "request" : {

            "type" : "text",

            "fields" : {

              "raw" : {

                "type" : "keyword"

              }

            }

          }

The reason why the "index": "not_analyzed" also removed because ES6.5 does not support this property of ...... checked the data seems to support only true and false. I changed the start type is text, but if raw set text can not seem to search. Now after this template set, our property is only used for fuzzy search, as the figure below, I will use logtype to search out a lot of irrelevant search results

 

 

And if the logtype.raw will precisely match. Personally, I guess this is the keyword effectiveness, not sure, I am not a specialized ES, so there is an error please correct me 

 

 

 

After modifying a template, you can successfully create log indice day, and the successful resolution of the issue 1

 

2. visualization and dashboard changes

This is speculation, after the upgrade we saved search, visualization dashboard there are almost ineffective. At that time only say that collapse, then slowly investigation found that the main problem lies in the search

visualization is based on the search, and then in turn is based on the visualization dashboard, so the search will repair fix everything

The problem is that every time are prompted to enter the next map

Saved object is missing

Could not locate that index-pattern (id:****)

click here to re-create it 

In fact, this problem before upgrading from 5.2 to 5.6 when it came once, when the solution is to re-create the index to delete a restored

The problem is still there delete N times

He gradually began to study the problem found in the search settings in

Settings where there is one "index": "logstash- *"

Try a bit of stick index ID into the problem resolved smoothly

Suspected previous version can be used to link the name index and search index, 6.5 does not support the case. Of course there would be any more likely to be the underlying API index save something, but because I did not find, so only use this method to solve the stupid

一键修复的话只需要导出所有的东西,然后find and replace

 

3 Kibana白屏

此问题至今无解,依旧在等AWS的回复,个人怀疑是升级过程中出现问题,在电话里我给他们的建议是尝试rollback然后重新升级一下试试

 

Guess you like

Origin www.cnblogs.com/ldgend/p/10953064.html