Duplicated index with different settings


(Mike W) #1

I would like to have two indexes, one with timestamp field defined and one without it. Is it possible?


(Joe Fleming) #2

If you are using two different indices, then they can have two different mappings, that's not a problem. If you are trying to query them together though, and you are using the timestamp field, the index without that field simply won't return any hits.


(Mike W) #3

But how do I create one under same name? Currently my 'little hack' is by creating two indexes by using different range of characters inside Kibana GUI.

Lets say ES stores data as my_little_db_test. Then, in Kibana, I create index by using two wildcards:

  1. my_little_db_test
  2. my_little_db_te*

But there must be a 'cleaner' way of doing this


(Joe Fleming) #4

Ah, I get what you're saying. If you're querying the same data from effectively the indices, I don't understand why you need a version of the data with and without a time field. I know there are some subtle differences in the way Kibana handles non-timeseries data, but is there something you can't do when you pick a time field on the index pattern?


(Mike W) #5

Lets just say I have no choice :slight_smile: