Kibana: Your Kibana index is out of date, reset it or use the X-Pack upgrade assistant.

[Borja13792]

Good morning,

I am trying to mount the kibana 7.00 but I get this error, there is some way to fix it.

plugin:[email protected] | Your Kibana index is out of date, reset it or use the X-Pack upgrade assistant.

AND with package.json this error:

This version of Kibana requires Elasticsearch v7.0.0-alpha1-SNAPSHOT

Thanks

[tylersmalley]

The Kibana index you are using was created with a version prior to 6.0 and needs to be migrated.

There are a few different options:

• Delete the Kibana index and allow one to be re-created (you will lose all visualization, dashboards, saved searches, etc)
• Manually migrate the index https://www.elastic.co/guide/en/kibana/6.0/migrating-6.0-index.html
• Install version 5.6, with X-Pack and use the Upgrade Assistant

[cdaringe]

in our docker setup (w/ ~no customization), we are running 6.x of ES & Kibana.

we tried:

• deleting the index and rebooting the things, yet the error persists
• doing the manual migration, yet the error persists

we are mounting our data--our other indices are picked by ES w/out issue. any idea?

[cdaringe]

in fact, or ES shows the new index and alias are zen, too:

$ curl -XGET 'localhost:9200/_cat/indices?v&pretty' | grep kib

green  open   .kibana-6                                                     qC75ww_-QTO6f8FmnPPmQg   1   0         81            0       50kb           50kb

$ curl -XGET 'localhost:9200/_cat/aliases?v&pretty' | grep kib
.kibana                                             .kibana-6                                                     -      -             -

yet kibana 6.0.0 still complains 🤔

[cdaringe]

sorry to keep spamming, but... more info:

elasticsearch_1  | [2017-11-21T21:44:40,962][INFO ][o.e.c.m.MetaDataCreateIndexService] [hyEjDLC] [.kibana-6] creating index, cause [api], templates [not_analyzed_template], shards [1]/[0], mappings [_default_, doc]
elasticsearch_1  | [2017-11-21T21:44:41,092][INFO ][o.e.c.r.a.AllocationService] [hyEjDLC] Cluster health status changed from [YELLOW] to [GREEN] (reason: [shards started [[.kibana-6][0]] ...]).
kibana_1         | {"type":"log","@timestamp":"2017-11-21T21:44:41Z","tags":["status","plugin:[email protected]","error"],"pid":1,"state":"red","message":"Status changed from yellow to red - Your Kibana index is out of date, reset it or use the X-Pack upgradeassistant.","prevState":"yellow","prevMsg":"No existing Kibana index found"}

• kibana inits a .kibana-6 index to ES
• ES boots it successfully
• kibana complains about it

[cdaringe]

we did some debugging:

kibana decides if your indicies have been migrated by testing the length of currentTypes, right here.

for instance, here are our currentTypes pre migration:

[ 'config',
  'visualization',
  'server',
  'timelion-sheet',
  'dashboard',
  '_default_',
  'index-pattern',
  'search',
  'graph-workspace',
  'url' ]

after running the migration process, our currentTypes yield:

[ 'doc', '_default_' ]

it would seem that _default_ is probably OK to be present? we dont know enough about ES internals to comment. however, it's clear that either:

• this check is invalid, or,
• ES mapping creation is invalid.

hope this helps!

[cdaringe]

@tylersmalley, just following up. as described above, there is some sort breach-of-contract between ES & Kibana. It is unclear if ES or Kibana is the offender.

this issue was started by a user using kibana 7.x, but in fact, we experience this issue w/ 6.x kibana and 6.x ES. let me know if I should move it to a separate issue or not.

[tylersmalley]

Thanks @cdahlqvist, will review your PR shortly. Do you have an index template which is adding _default_ to the new index?

[cdaringe]

@tylersmalley, it would seem so, from kibana itself.

$ curl http://localhost:9200/_template | jq '.["default_kibana_template"]["mappings"]["_default_"]'
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   866  100   866    0     0  73714      0 --:--:-- --:--:-- --:--:-- 78727
{
  "dynamic_templates": [
    {
      "strings": {
        "match_mapping_type": "string",
        "mapping": {
          "fields": {
            "keyword": {
              "ignore_above": 256,
              "type": "keyword"
            }
          },
          "type": "text"
        }
      }
    }
  ]
}

[tylersmalley]

@cdaringe, there would have been another template which set the _default_ type.

https://github.com/elastic/kibana/pull/15432 has been merged, so closing this issue.

[cdahlqvist]

@tylersmalley I do not understand why I am included in this thread.

[tylersmalley]

@cdahlqvist, apologies - It was a mention auto-complete fail.

[rockmelodies]

我也是这个问题，该怎么解决呢？
I am the same question, how should I solve it?

[tylersmalley]

@rockmelodies read my first response here.

Most common solution is to manually migrate the index following these steps

[osehgol]

Deleting Kibana index worked for me (CAUTION: I had a brand new install),

curl -XDELETE http://localhost:9200/.kibana

I got it from https://stackoverflow.com/questions/31201051/elasticsearch-is-still-initializing-the-kibana-index

[ghost]

Deleting Kibana index worked for me (CAUTION: I had a brand new install),

curl -XDELETE http://localhost:9200/.kibana

[ghost]

Can anyone provide any explanation to why doing this solved the issue?

[jebrage]

After trying @osehgol's solution I got a bad gateway error at my kibana site. But it worked once I restarted all of the elk services.

[LiosWong]

Deleting Kibana index worked for me (CAUTION: I had a brand new install),

curl -XDELETE http://localhost:9200/.kibana

I got it from https://stackoverflow.com/questions/31201051/elasticsearch-is-still-initializing-the-kibana-index

thanks!