How do you improve profile of a match_all query that takes 20s?

[TealFawn (Jef Statham)]

Doing the basic match all query on one index takes a cumulative time of 20s or 11s from one shard. Where should I look to improve search performance on the most basic of searches I can do.

Query:
{
"query":{
"match_all" : {}
}
}

[xavierfacq]

Hi, You should provide more informations about your cluster, indices, shards and configurations.

[TealFawn (Jef Statham)]

Not sure what exactly to provide cluster_state.json is an ~8mb file.

Here's the overview though

It's also worth noting with the majority of the indices have 5 shards with 1 replica.