We are having trouble with some slow aggregation queries and troubleshooting has guided me in the direction of a possible low filter cache.
Based on this gist, it seems like this was available from different APIs earlier:
I have tried these endpoints, without finding anything related to "filter":
_nodes/stats?all
_cluster/stats?all
Have also looked in marvel, without finding any reference to the filter cache.
Is it no longer possible to monitor the size of the filter cache?