[akireu]

On a side note, Prometheus seems to be built for bloat. AFAIK, it isn't even designed to consume metrics other than from apps linked to its client library. It's like a microservice, but with the footprint of an operating system.

[hughrr]

It’s a database engine not a microservice and needs to be treated along the lines of postgresql etc.

[yencabulator]

  podman run postgres:14 du -sh /usr/lib/postgresql
  24M /usr/lib/postgresql

This includes the binary.

[hughrr]

Try running big instances in containers.

Our smallest is 32 cores and 1TB of RAM.

[momothereal]

> AFAIK, it isn't even designed to consume metrics other than from apps linked to its client library.

Could you elaborate? I use Prometheus to scrape from an HTTP endpoint in various Pods in Kubernetes, so the service discovery is pretty useful to me.

I could see the Kubernetes & the other SDs split out of the core binary if default size is really an issue. Or are you talking about something else?

[akireu]

I'm talking about the way I'm expected to provide metrics for my apps. Rather than exporting free-form JSON and then scripting Prometheus to understand it, I'm expected to use a custom client library to export the metrics. As for Kubernetes, you can only use it with Prometheus because of not insignificant amount of work on both sides. Basically, the latter is designed for vendor lock-in.

[slimsag]

What a bizarre claim.

Prometheus scrapes the same text format as OpenMetrics 1.0 and over 700 public exporters use this format, and there are TONS of other non-Prometheus software that consume the exact same text format. Prometheus's biggest competitor, Datadog (which is not open source mind you), consumes it too. I think even Grafana consumes it directly. It's becoming an IETF standard[0].

Would I have preferred JSON over a custom text format like this? Yeah. But to claim an open source project like Prometheus with effectively no business at all is using a text format like this to have vendor lock-in? That's quite a stretch.

[0] https://github.com/OpenObservability/OpenMetrics/blob/main/s...

[morelisp]

> Prometheus scrapes the same text format as OpenMetrics 1.0

I find the GP's claims weird - I've written a relative ton of collectors, exporters, and translators and the format is pretty OK, not worse than most that came before it and better than lots - but I think this relationship is backwards. Prometheus "scrapes OpenMetrics" because OpenMetrics was formal documentation of what Prometheus was already doing for years.

I would not have preferred JSON. That an exposed metric is also a query is also pretty close to a schematic definition is nice.

[akireu]

I apologize for my mistake, then. My understanding was based on reading the Prometheus docs on making exporters alone - something I needed urgently for a job.

[slimsag]

no apology needed, I am sorry the world has a culture of mistakes being a bad thing.

[jrockway]

Include the client library if you want, but the wire format is ridiculously simple. I'll implement it from memory in a HN comment.

    http.HandleFunc("/metrics", func(w http.ResponseWriter, req *http.Request) {
        w.WriteHeader(http.StatusOK)
        w.Header().Add("content-type", "text/plain")
        w.Write([]byte("# HELP foo_bar The numbers of foos barred.\n# TYPE foo_bar counter\nfoo_bar 42\n"))
    })

The client library is largely to keep track of running counters (and gauges, histograms, etc.), with a small amount of code to actually report those metrics when scraped. It's a very simple format.

[morelisp]

A good example of it looking simple but having annoying corner cases.

The content type MUST be: application/openmetrics-text; version=1.0.0; charset=utf-8

[gempir]

Prometheus follows the OpenMetrics standard I'm not sure what you find propietary about that or specific to prometheus.

https://github.com/OpenObservability/OpenMetrics/blob/main/s...

[momothereal]

To be precise it was the other way around. OpenMetrics is a standardization effort for the format Prometheus made up.

However Prometheus was designed before JSON was standardized itself, so I'm just glad they didn't choose XML!

[lokar]

And the Prometheus format is a copy of the "varz" format

[SuperQue]

Sorta, I would say it's an evolution of varz.

IIRC (it's been almost a decade since I used varz), having multiple label values would be a map of maps in varz. It got quite ugly if you wanted to have a number of dimensions.

[momothereal]

Ah ok, I see what you mean.

The other commenters have pointed out that it _is_ based on another open standard, but admittedly one less common than say, JSON. So you'll generally have to implement your own metrics producer or use a client library, that's true.

However it's also a dead simple format and you can probably implement it with a for-loop or a shell script.

[SuperQue]

Prometheus supported a JSON representation in the beginning. It was deprecated and removed before 1.0. The current exposition format was created because it cut CPU and memory for scraping metrics in half.

JSON, especially free-form JSON, is not a good format for efficient metrics monitoring.

[NikolaeVarius]

The prometheus format is literally just a text page. It dead simple to implement

[rixed]

The design consideration was not that it had to be simple to implement. It's that it had to be easy to parse by a human during an outage when nothing else works.

[SuperQue]

It's a little bit of both. Simple to implement, simple (and fast/cheap) to parse, and human readable.

[morelisp]

There are a frustrating number of fundamental corner cases due variance to floating point text formats, and slightly more in the descriptor if you also need that. It's simple to implement an expositor for a limited set of cases. As usual, it's much more difficult to parse what you actually find in the world.

[morelisp]

Maybe I'm being downvoted for insufficient examples? OK, here's a big one:

OpenMetrics's production rule for the format says:

    labels = "{" [label *(COMMA label)] "}"

And yet, the Prometheus Java client library exports a trailing comma with no subsequent label.

As for fp, I've seen parsers break on `e` v. `E`, and `NaN`/`Inf`/ vs. `nan`/`inf`. The latest IETF draft even has a comment,

     ; Not 100% sure this captures all float corner cases.

Control characters are allowed in descriptors and label values and no normalization form is specified.

[SuperQue]

Yea, there still some corner cases and implementation bugs out there. We spent months deliberating how to deal with some of these. Because the base libraries in some languages just don't produce string output from IEEE 754 the same way.

IIRC, Java is different from Python is different from Go. So, really, this is a standardization in languages problem. We tried to work around these as best we could in the OM format.

[atmosx]

> It's like a microservice, but with the footprint of an operating system.

What? It's 100mb vs 25mb, like 75MB of data more... Who cares about a binary being 75MB larger in 2022?

[jasonjayr]

This is a frustrating position.

Everything is layers. We build things on top of other things. If every layer had that attitude, then the bloat would be enormous. It's already getting there.

We should praise judicious effort into optimizing any of the resources used in the systems build, at every layer.

[atmosx]

I understand where you're getting at. IMO you're barking at the wrong tree: the problem with bloatware will not be solved by prometheus shipping a lighter statically linked binary.