mimir VS m3

It consists of 6 separate components to run, plus 4 additional optional components [1]. And I don't even count 5 additional caches [2]. How is that "simple and easy"?

[1] https://grafana.com/docs/mimir/latest/references/architectur...

[2] https://github.com/grafana/mimir/blob/9a52522a05f25d6d6bc2e8...

Yes, but also no. The Prometheus ecosystem already has two FOSS time-series databases that are complementary to Prometheus itself. Thanos and Mimir. Not to mention M3db, developed at Uber, and Cortex, then ancestor of Mimir. There's a bunch of others I won't mention as it would take too long.

I used the one in their tutorial.

Grafana Agent can collect and stream data via Prometheus remote write protocol. Then you can either use Prometheus or Mimir. to receive the data at a central location. It depends on how many millions of series you need to collect.

Long term storage all depends on your needs and sophistication. I use Thanos for our system since it has an extremely flexible scaling system. But there is also Grafana Mimir. They're both similar in that they use Prometheus TSDB format as part of the underlying storage. One nice Thanos advantage is that it does do downsampling in addition to being able to store raw metric data for a long time. It will auto-select downsampled data to make requests faster.

Grafana Mimir - scalable storage for Prometheus

Depending on how you want to set things up, you can use Thanos or Mimir to create the single-pane-of-glass view of your data.

M3DB.

> I can't find any other open source time series database except Mimir/Cortex which allows this much scale (clustering options in their open source version)

The following open source time series databases also can scale horizontally to many nodes:

- Thanos - https://github.com/thanos-io/thanos/

- M3 - https://github.com/m3db/m3

- Cluster version of VictoriaMetrics - https://docs.victoriametrics.com/Cluster-VictoriaMetrics.htm... (I'm CTO at VictoriaMetrics)

> Can we use Prometheus/Mimir as general purpose time series database?

This depends on what do you mean under "general purpose time series database". Prometheus/Mimir are optimized for storing (timestamp, value) series where timestamp is a unix timestamp in milliseconds and value is a floating-point number. Each series has a name and can have arbitrary set of additional (label=value) labels. Prometheus/Mimir aren't optimized for storing and processing series of other value types such as strings (aka logs) and complex datastructures (aka events and traces).

So, if you need storing time series with floating-point values, then Prometheus/Mimir may be a good fit. Otherwise take a look at ClickHouse [1] - it can efficiently store and process time series with values of arbitrary types.

[1] https://clickhouse.com/