Hi,
we have an opensearch cluster with 6 nodes and 16TB docs usage.
There is an option ‘index.mapping.total_fields.limit’ which limits the fields in an index. As far as I have researched, more fields mean higher memory usage and increased CPU load.
We would like to add ~500-1000 unique additional fields to all indices. Is it a problem to increase the value from 1000 to, for example, 4000? I know that indexing more fields requires more CPU usage during indexing, and each field consumes more memory.
Is the adjustment of the value really noticeable?
Regards,
Thomas