VictoriaMetrics VS influxdb_iox

VictoriaMetrics CTO here.

The referred library is the official OpenTelemetry package for reading metrics in Go language [1] - more details are available at [2].

Note that we at VictoriaMetrics like the idea of unified observability standard like OpenTelemetry. The issue is in the current otel implementation. It is too bloated and very inefficient. This contradicts to our experience with observability cases, which need very optimized format for metrics' transfer in order to reduce costs on CPU and network traffic needed to transfer and process these metrics.

VictoriaMetrics continues investing in OpenTelemetry by providing integration docs [3] and improving the existing functionality for otel metrics' ingestion [4].

[1] https://github.com/open-telemetry/opentelemetry-proto-go

[2] https://github.com/VictoriaMetrics/VictoriaMetrics/pull/2570...

[3] https://docs.victoriametrics.com/guides/getting-started-with...

[4] https://github.com/VictoriaMetrics/VictoriaMetrics/issues/60...

Victoria Metrics

VictoriaMetrics is primarily a time-series database designed for efficiently storing and querying time-series data. It is often used as a back-end data store for time-series data generated by monitoring systems like Prometheus. VictoriaMetrics excels at handling large volumes of time-series data, offering efficient storage and query capabilities.

Not sure I follow since there are very competitive tools written in Go such as https://victoriametrics.com for an example in this space.

Did you try VictoriaMetrics [1] and vmagent [2]? It is a single self-contained binary without external dependencies. It requires relatively low amounts of CPU, RAM, disk space and disk IO, and it runs on ARM.

[1] https://github.com/VictoriaMetrics/VictoriaMetrics/

[2] https://docs.victoriametrics.com/vmagent.html

https://github.com/VictoriaMetrics/VictoriaMetrics#cardinali...

If I understanding correctly, it deal with high cardinality by dropping data, the operators need to monitor for this and adjust their data to lower the cardinality.

VictoriaMetrics

You shouldn't unless you want to use the new open source standard for telemetry. You won't benefit from simplicity or performance improvements. It would be quite the opposite. You can check what is the actual cost of open telemetry adoption here [0]

But if you ever decide to go this path - VictoriaMetrics supports OpenTelemetry protocol for metrics [1]

[0] https://github.com/VictoriaMetrics/VictoriaMetrics/pull/2570

[1] https://docs.victoriametrics.com/Single-server-VictoriaMetri...

Watch out for the AGPL minio <https://github.com/metrico/iox-community/blob/155a14bb5e8e32...> the almost certainly AGPL grafana <https://github.com/grafana/grafana/blob/v10.1.1/LICENSE> and always eye anyone who uses :latest images with healthy suspicion

That said, influx_iox itself appears to be Apache 2 (and/or MIT?) https://github.com/influxdata/influxdb_iox/blob/main/LICENSE...

have you looked at https://github.com/influxdata/influxdb_iox ? that's where the development for the new version is done.

As I understand, InfluxDB 3 is just a re-branding of InfluxDB IOx. Then its' performance can be not very good comparing to Prometheus-like systems.

InfluxDB iox

Don't know how much is open or closed, but they were doing some development in the open: https://github.com/influxdata/influxdb_iox

Just want to say congratulations to the team!

2 years and 9,500+ commits is a hell of a feat.

https://github.com/influxdata/influxdb_iox

Already is: https://github.com/influxdata/influxdb_iox Just still a work in progress.

It looks like this crate is the workaround for that. But there's a PR on SQLX opened a couple days ago that will fix the issue.

You should check out polars, datafusion, influxdb iox and databend, all written in native Rust and powered by the Apache Arrow format. Polars in particular is pretty dam fast and has bindings for Python.