veneur VS opentelemetry-proto

This was the idea behind Stripe's Veneur project - spans, logs, and metrics all in the same format, "automatically" rolling up cardinality as needed - which I thought was cool but also that it would be very hard to get non-SRE developers on board with when I saw a talk about it a few years ago.

https://github.com/stripe/veneur

One pain point with Prometheus is that is has relatively weak support for quantiles, histograms, and sets[1]:

- Histograms require manually specifying the distribution of your data, which is time-consuming, lossy, and can introduce significant error bands around your quantile estimates.

- Quantiles calculated via the Prometheus "summary" feature are specific to a given host, and not aggregatable, which is almost never what you want (you normally want to see e.g. the 95th percentile value of request latency for all servers of a given type, or all servers within a region). Quantiles can be calculated from histograms instead, but that requires a well-specified histogram and can be expensive at query time.

- As far as I know, Prometheus doesn't have any explicit support for unique sets. You can compute this at query time, but persisting and then querying high-cardinality data in this way is expensive.

Understanding the distribution of your data (rather than just averages) is arguably the most important feature you want from a monitoring dashboard, so the weak support for quantiles is very limiting.

Veneur[2] addresses these use-cases for applications that use DogStatsD[3] by using clever data structures for approximate histograms[4] and approximate sets[5], but I believe its integration with Prometheus is limited and currently only one-way - there is a CLI app to poll Prometheus metrics and push them into Veneur, but there's no output sink for Veneur to write to Prometheus (or expose metrics for a Prometheus instance to poll).

It would be extremely useful to have something similar for Prometheus, either by integrating with Veneur or implementing those data structures as an extension to Prometheus.

[1] https://prometheus.io/docs/practices/histograms/

[2] https://github.com/stripe/veneur

[3] https://docs.datadoghq.com/developers/dogstatsd/

[4] https://github.com/stripe/veneur#approximate-histograms

[5] https://github.com/stripe/veneur#approximate-sets

Maybe, you are asking yourself: "But I already had instrumented my applications with vendor-specific libraries and I'm using their agents and monitoring tools, why should I change to OpenTelemetry?". The answer is: maybe you're right and I don't want to encourage you to update the way how you are doing observability in your applications, that's a hard and complex task. But, if you are starting from scratch or you are not happy with your current observability infrastructure, OpenTelemetry is the best choice, independently of the backend telemetry tool that you are using. I would like to invite you to take a look at the number of exporters available in the collector contrib section, if your backend tracing tool is not there, probably it's already using the Open Telemetry Protocol (OTLP) and you will be able to use the core collector. Otherwise, you should consider changing your backend telemetry tool or contributing to the project creating a new exporter.

Here's the example payloads for OTLP over JSON and example of how to ingest them: https://github.com/open-telemetry/opentelemetry-proto/tree/m...

An OTLP receiver can receive data via gRPC or HTTP using the OTLP format. There are advanced configurations that you can enable via the YAML file.

Cerbos fully transitioned from OpenCensus to OpenTelemetry, a move that significantly boosts our metrics and tracing capabilities. This shift allows for more efficient integration with a variety of observability products supporting the OpenTelemetry protocol (OTLP) but also offers the flexibility to use push metrics and fine-tune trace sampling. With this update, configuration through the tracing block in Cerbos files is deprecated in favor of using OpenTelemetry environment variables.

Networking is Easy (Really!) Since you install the agent directly into the environment where you are running your application, there is no complex networking. When developing in localMode, the agent listens on the common OpenTelemetry Line Protocol (OTLP) on ports 4317 & 4318 automatically.

Oh nice, thank you (and also solumos) for the links! It looks like oteps/pull/171 (merged June 2023) expanded and superseded the opentelemetry-proto/pull/346 PR (closed Jul 2022) [0]. The former resulted in merging OpenTelemetry Enhancement Proposal 156 [1], with some interesting results especially for 'Phase 2' where they implemented columnar storage end-to-end (see the Validation section [2]):

* For univariate time series, OTel Arrow is 2 to 2.5 better in terms of bandwidth reduction ... and the end-to-end speed is 3.1 to 11.2 times faster

* For multivariate time series, OTel Arrow is 3 to 7 times better in terms of bandwidth reduction ... Phase 2 has [not yet] been .. estimated but similar results are expected.

* For logs, OTel Arrow is 1.6 to 2 times better in terms of bandwidth reduction ... and the end-to-end speed is 2.3 to 4.86 times faster

* For traces, OTel Arrow is 1.7 to 2.8 times better in terms of bandwidth reduction ... and the end-to-end speed is 3.37 to 6.16 times faster

[0]: https://github.com/open-telemetry/opentelemetry-proto/pull/3...

[1]: https://github.com/open-telemetry/oteps/blob/main/text/0156-...

[2]: https://github.com/open-telemetry/oteps/blob/main/text/0156-...

We then modified the benchmark to encode our example data which is an opentelemetry trace data.