Which TIME SERIES DATABASE to choose for processing and storing 100 million points per day?

Question

big data

To collect metrics from my projects, I used InfluxDB and at first everything was fine, but after increasing the number of different metrics, InfluxDB began to fall too often. It turned out that when I try to do some calculations for the day, the database consumes too many resources and as a result, it simply does not have enough of them to make these calculations. The increase in resources instilled an improvement, but only a temporary one, he still demanded more and more.
InfluxDB is good, but for a small amount of data. Perhaps InfluxEnterprise would fit in well with my problem, but I would like to use only open source solutions.
The first time I chose Time-Series I followed the trend db-engines.com/en/ranking/time+series+dbms. Now I see that the trend did not suit me and now apparently I can’t do without load testing of all existing TSDBs). Therefore, I want to turn to you to save time when choosing the best tool.
After reading all of the above, please answer 2 simple questions:

• What modern time-series DB do you use in production?
  
• What can help me in solving this problem?
  

Now I'm thinking of looking at the following TSDBs:
OpenTSDB Axibase
Chronix
Cube
Heroic
InfluxDB
Kairosdb
Newts
Prometheus
TrailDB
Riak-TS Riak TS
Akumuli Akumuli
Rhombus
Dalmatiner
Blueflood
Timely Timely

Answer 1

Did you watch this? https://habrahabr.ru/company/yandex/blog/303282/