feat: produce OpenTelemetry traces with hs-opentelemetry

[develop7]

This PR introduces producing OpenTelemetry traces containing, among others, metrics same as in ServerTiming header from before.

TODO:

• build with Nix as well (for now Stack only)
• make an example of exporting log messages hs-opentelemetry doesn't support logging, per Logging roadmap iand675/hs-opentelemetry#100
• export actual error
• make getTracer available globally: we're interested in using as many different spans as it makes sense, so getTracer should be available everywhere, as described in hs-opentelemetry-sdk's README
• make use of hs-opentelemetry-wai middleware
• look into failing Windows builds hs-opentelemetry-sdk depends on unix, tracking in Windows support iand675/hs-opentelemetry#109
• fix "memory test" failures (https://github.com/PostgREST/postgrest/actions/runs/7888534966/job/21526302784?pr=3140)
• fix dependency issues on 9.8 (https://github.com/PostgREST/postgrest/actions/runs/7888534966/job/21526300453?pr=3140)

Running:

I sort of gave up deploying and configuring all the moving bits locally, so you'd need to create the honeycomb.io account for this one (or ask me for the invite). After that, it's quite straightforward:

1. Build PostgREST executable with stack build, and get its path with stack exec -- which postgrest
2. Get a PostgreSQL server running (e.g. run nix-shell, then postgrest-with-postgresql-15 -- cat)
3. Do the JWT dance (generate the JWT secret and encode the role into a token, e.g. postgrest_test_anonymous for the example DB from the above example)
4. Run PostgREST server with

   OTEL_EXPORTER_OTLP_ENDPOINT='https://api.honeycomb.io/' \
   OTEL_EXPORTER_OTLP_HEADERS="x-honeycomb-team=<honeycomb_api_key>"  \
   OTEL_SERVICE_NAME='PostgREST' OTEL_LOG_LEVEL='debug' OTEL_TRACES_SAMPLER='always_on' \
   PGRST_DB_URI='<postgresql_server_url>'  PGRST_JWT_SECRET='<jwt_secret>'  \
   /path/to/compiled/bin/postgrest
   

5. request some data and check the honeycomb dashboard for the traces:

[steve-chavez]

Awesome work! 🔥 🔥

I sort of gave up deploying and configuring all the moving bits locally, so you'd need to create the honeycomb.io account for this one

Found this Nix flake that contains an OTel GUI: https://flakestry.dev/flake/github/FriendsOfOpenTelemetry/opentelemetry-nix/1.0.1

I'll try to integrate that once the PR is ready for review.

[develop7]

The recent problem I'm seemingly stuck with is hs-opentelemetry is using UnliftIO, which seems not quite composable with our (implicit, correct?) monad stack. So the deeper into the call stack the instrumented code is (the one I'm trying to wrap with inSpan), the more ridiculously complex it should be changed to be instrumented, i.e. https://github.com/PostgREST/postgrest/pull/3140/files#diff-5de3ff2b2d013b33dccece6ead9aeb61feffeb0fbd6e38779750511394cf9701R156-R157, up to the point I have no idea how to proceed further (e.g. wrapping App.handleRequests cases with their own spans, which is semantically correct)

There's a more straightforward MonadIO-involving opentelemetry library, with less activity and quite different approach to the telemetry data export (GHC eventlog → file/pipe by the GHC runtime). It looks less invasive approach, refactoring-wise, but requires more hoops to jump to actually deliver traces to Honeycomb/Lightstep/whatnot (pull eventlog → convert it to zipkin/jaeger/b3 → upload somewhere for analysis).

It also seems to boil down to the conceptual choice between online and offline traces' delivery-wise, or push and pull model.

@steve-chavez @wolfgangwalther @laurenceisla what do you think guys?

[steve-chavez]

@develop7 Would vault help? It was introduced on #1988, I recall it helped with IORef handling.

It's still used on

postgrest/src/PostgREST/Auth.hs

Lines 160 to 165 in d2fb67f

	jwtDurKey :: Vault.Key Double
	jwtDurKey = unsafePerformIO Vault.newKey
	{-# NOINLINE jwtDurKey #-}
	getJwtDur :: Wai.Request -> Maybe Double
	getJwtDur = Vault.lookup jwtDurKey . Wai.vault

I'm still not that familiar with OTel but the basic idea I had was to store these traces on AppState and export them async.

[steve-chavez]

@develop7 Recently merged #3213, which logs schema cache stats to stderr. Perhaps that can be used for introductory OTel integration instead? It might be easier since the scache stats are already in IO space.

[develop7]

Would vault help?

hs-opentelemetry is using it already

basic idea I had was to store these traces on AppState and export them async

Not only that, you want traces in tests too, for one.

The good news is hs-opentelemetry-utils-exceptions seems to be just what we need, let me try it.

Perhaps that can be used for introductory OTel integration instead?

Good call @steve-chavez, thank you for the suggestion. Will try too.

[develop7]

it works!

[steve-chavez]

Since now we have an observer function and Observation module

postgrest/src/PostgREST/App.hs

Lines 170 to 172 in 229bc77

	handleRequest :: AuthResult -> AppConfig -> AppState.AppState -> Bool -> Bool -> PgVersion -> ApiRequest -> SchemaCache ->
	Maybe Double -> Maybe Double -> (Observation -> IO ()) -> Handler IO Wai.Response
	handleRequest AuthResult{..} conf appState authenticated prepared pgVer apiReq@ApiRequest{..} sCache jwtTime parseTime observer =

postgrest/src/PostgREST/Observation.hs

Lines 15 to 18 in 229bc77

	data Observation
	= AdminStartObs (Maybe Int)
	| AppStartObs ByteString
	| AppServerPortObs NS.PortNumber

Perhaps we can add some observations for the timings?

Also the Logger is now used like:

postgrest/src/PostgREST/Logger.hs

Lines 53 to 54 in 7c6c056

	logObservation :: LoggerState -> Observation -> IO ()
	logObservation loggerState obs = logWithZTime loggerState $ observationMessage obs

postgrest/src/PostgREST/CLI.hs

Line 50 in 7c6c056

CmdRun -> App.run appState (Logger.logObservation loggerState))

For OTel, maybe the following would make sense:

otelState <- Otel.init

App.run appState (Logger.logObservation loggerState >> OTel.tracer otelState)) 

[develop7]

Perhaps we can add some observations for the timings?

Agreed, server timings definitely belong there.

[develop7]

Okay, the PR is in the cooking for long enough, let's pull the plug and start small. Let's have it reviewed while I'm fixing the remaining CI failures.

[wolfgangwalther]

I've been working hard lately to get rid of non-hackage dependencies and would not like to introduce them again. Why do we need a fork here?

[develop7]

The original hs-opentelemetry depends on unix, which fails to build on Windows. But there are only two functions it uses from unix, getProcessID and getEffectiveUserID. The former is provided by unix-compat, the latter isn't. So the fork replaces unix dependency with unix-compat and removes the collection of "effective username" attribute, making it build on Windows here and now.

[wolfgangwalther]

I see. Do you plan to upstream your fixes into hs-opentelemetry itself?

[develop7]

I absolutely am, and I have no intent to maintain the fork more than I need to. Windows support is tracked upstream at iand675/hs-opentelemetry#109.

[wolfgangwalther]

hs-opentelemetry is, according to the repo, in alpha state. According to the TODO list above, the issue tracker and the repo description, it does not support:

• GHC 9.8.x
• Windows
• Metrics or Logging

I don't think we depend on this in the current state. And we should certainly not depend on an even-less-maintained fork of the same.

So to go forward here, there needs to be some effort put into the upstream package first, to make it usable for us.

[develop7]

A status update:

• GHC 9.8: hs-opentelemetry-sdk doesn't build against 9.8 because of hs-opentelemetry-exporter-otlp → proto-lens chain. Given the upstream of the latter being bit unresponsive for the suggestions to bump upper bounds, I've managed to make the latter build for 9.8 in develop7/proto-lens@985290f, but haven't figured out how to pick it up to the project since it depends on the google's protobuf compiler installed and the protobuf's source checked out. Another approach is to not use hs-o-sdk and hs-o-e-otlp altogether, which I probably should've tried way before.

[wolfgangwalther]

• GHC 9.8: hs-opentelemetry-sdk doesn't build against 9.8 because of hs-opentelemetry-exporter-otlp → proto-lens chain. Given the upstream of the latter being bit unresponsive for the suggestions to bump upper bounds, I've managed to make the latter build for 9.8 in develop7/proto-lens@985290f,

Hm. I looked at your fork. It depends on support for GHC 9.8 in ghc-source-gen. This repo has a PR, which just was updated 3 days ago. I wouldn't call that "unresponsive", yet. Once ghc-source-gen is GHC 9.8 compatible, you could open a PR to update bounds in proto-lens itself. But since the last release for GHC 9.6 support was in December... I would not expect this to take too long to get responded to. It certainly doesn't look like it's unmaintained.

I guess for GHC 9.8 support it's just a matter of time.

What about the other issues mentioned above? Were you able to make progress on those?

[mkleczek]

The recent problem I'm seemingly stuck with is hs-opentelemetry is using UnliftIO, which seems not quite composable with our (implicit, correct?) monad stack. So the deeper into the call stack the instrumented code is (the one I'm trying to wrap with inSpan), the more ridiculously complex it should be changed to be instrumented, i.e. https://github.com/PostgREST/postgrest/pull/3140/files#diff-5de3ff2b2d013b33dccece6ead9aeb61feffeb0fbd6e38779750511394cf9701R156-R157, up to the point I have no idea how to proceed further (e.g. wrapping App.handleRequests cases with their own spans, which is semantically correct)

There's a more straightforward MonadIO-involving opentelemetry library, with less activity and quite different approach to the telemetry data export (GHC eventlog → file/pipe by the GHC runtime). It looks less invasive approach, refactoring-wise, but requires more hoops to jump to actually deliver traces to Honeycomb/Lightstep/whatnot (pull eventlog → convert it to zipkin/jaeger/b3 → upload somewhere for analysis).

It also seems to boil down to the conceptual choice between online and offline traces' delivery-wise, or push and pull model.

@steve-chavez @wolfgangwalther @laurenceisla what do you think guys?

In my prototype I actually played with replacing HASQL Session with an https://github.com/haskell-effectful/effectful based monad to make it extensible:

https://github.com/mkleczek/hasql-api/blob/master/src/Hasql/Api/Eff/Session.hs#L37

Using it in PostgREST required some mixins usage in Cabal:

29b946e#diff-eb6a76805a0bd3204e7abf68dcceb024912d0200dee7e4e9b9bce3040153f1e1R140

Some work was required in PostgREST startup/configuration code to set-up appropriate effect handlers and middlewares but the changes were quite well isolated.

At the end of the day I think basing your monad stack on an effect library (effectful, cleff etc.) is the way forward as it makes the solution highly extensible and configurable.

[develop7]

Update: rebased the PR against latest master, updated hs-opentelemetry (with the Windows support merged!) & asked hs-opentelemetry maintainers to cut a new release in iand675/hs-opentelemetry#154 so we don't have to depend on forks again.

[develop7]

Building hs-opentelemetry from source introduces a build-time dependency of protoc, a protobuf compiler, so that's why Cabal & Stack builds fail

[steve-chavez]

Hey @develop7!

QQ, this would only grant us otel traces for the JSON error responses right? It would not send any other logs and I think otel is meant to send these as well?

[develop7]

No explicit inSpan* calls means no traces altogether. https://hackage.haskell.org/package/hs-opentelemetry-instrumentation-auto could help with that, but it requires MonadUnliftIO and, I guess, adopting mtl style?

It would not send any other logs and I think otel is meant to send these as well?

No it won't; not at the moment — while OTel spec does have logs, hs-opentelemetry is yet to support them, as well as metrics.

[steve-chavez]

I still wonder if we could do this in a less invasive way 🤔. We added Prometheus metrics in a simple way by using our Observation module:

postgrest/src/PostgREST/AppState.hs

Line 133 in 2564b32

let observer = liftA2 (>>) (Logger.observationLogger loggerState configLogLevel) (Metrics.observationMetrics metricsState)

Ideally we would add an Otel module, add an Otel.traces function there and that would be it.

I'm not sure if this can be done entirely with hs-opentelemetry but in theory we would need a grpc dependency (format used by otel) plus an http client that does the work in a separate thread.

It seems better to maintain some of this functionality in-tree, specially since the dependency is not fully featured yet. That would also give us better control.

[develop7]

@steve-chavez this could be even more preferable since every observer invocation is accompanied by the tag, so I don't have to invent span labels. As long as observer is called in all/most interesting places, right?

I'm not sure if this can be done entirely with hs-opentelemetry

It seems so — I didn't sent any of these traces to honeycomb myself, so. I might only need to fiddle with the call stack so it won't have observer on top, but even that will do.

[develop7]

hs-opentelemetry maintainers released an update to 0.1.0.0 yesterday, and with that the protoc requirement is gone again and my windows-enabling contribution is in; that should also fix Nix builds as well since we don't need VCS dependencies again (well, for now)

[develop7]

Nix builds still fail because nixpkgs we're using do not have updated hs-opentelemetry-*. I'm on to update haskell-packages.nix accordingly.