MongoDB. Why are index queries so slow?

Question

MongoDB

Hello.
I ask for help in solving the following problem:
There is a collection of 26 million documents in size.
The structure of the document is fixed - pastebin.com/iBzW0Fkz
An index is built on the users field - db.conversations.getIndexes() - pastebin.com/xHecpw00
Queries on the users field are slow (more than 100 ms)
db.getCollection('conversations'). find({users: {$all: ["5942328", "9082468"]}}); EXPLAIN - pastebin.com/0C11Cr9F
db.getCollection('conversations').find({users: "9163099"}); EXPLAIN - pastebin.com/CmvuDf10
Questions:
1. Why are index queries so slow?
2. How to warm up (load into RAM) indexes after restarting the mongod server/process? In WiredTiger, the touch() command has been removed.
More information on collection and database:
db.stats() - pastebin.com/9JZF8ChQ
db.getCollection('conversations').stats() - pastebin.com/17yV4Fsi
db.conversations.getIndexes() - pastebin.com/xHecpw00
Server information:
lscpu - pastebin.com/k7wUE4gH
lshw -short - pastebin.com/w5XYuY3U
There is a suggestion that reading from the HDD is very slow and the solution would be to install an SSD, but we need to explicitly confirm this.

Answer 1

Try to select not the entire document:

db.conversations.find({users: "9163099"}, {_id: 1})
или
db.conversations.find({users: "9163099"}, {users: 1})

What does free -m show?
If I were you, I would buy a VPS in DO for half an hour, and check it - it costs a penny (5-30 rubles) and it’s a matter of a minute * (maybe the problem is not only in the HDD), if it’s a hetzner, then everything is clear.