I have wierd contradiction between the number of fileds in elasticsearch and the number of fields in the corresponding kibana index.
I have two indices in elasticsearch called: nul_check_1 and nul_check_2.
I check how many mapping there are in both of them by running:
curl -s -XGET localhost:9200/nul_check*/_mapping?pretty | grep type | wc -l
I get the same number if I run this command for each index and summing up the results.
When I check the kibana index that matches these indices, in "management" tab, I see less fields that what I saw in elasticsearch.
Why is this happening?
How can I assess how many fields there are? It's relevant because the cap of elasticsearch on 1000 fields by default.
Kibana stores a copy of the elasticsearch mappings because we attach kibana specific metadata to fields (field formatters, etc). If the elasticsearch mappings update, Kibana's mappings need to be "refreshed" to get the latest. That should give you the up to date number of fields.
Refreshing can be done on the index management page, in the top right there's a refresh icon.
Even after refreshing the number is still smaller.
I suspect the curl command is being too permissive. Are there multiple indices with the same mappings being returned? Are there any fields named "type"?
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.