Fill otel demo with more data required before kubecon #2127
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
marcelbirkner
commented
Mar 20, 2025
marcelbirkner
commented
- add more logs to product catalog service
- update deployment automation script
- update README how to create a new release
Also: * propagate the change from bool to float introduced in open-telemetry#1237 more consistently via proto definitions by differentiating between the GetFlag operation (which evaluates the probabilty and therefore returns a bool) and all other operations, which need to operate with a float value/probability directly. To that end, the Flag grpc message has been split into two new types, FlagEvaluationResult and FlagDefinition. * Rename the UpdateFlag operation to UpdateFlagProbability, as it actually only updates the enabled/probability value, but not the description or the name.
Instead of requiring a git release as the trigger for publishing container images, add a workflow_dispatch trigger and set a fixed version number.
* allow arbitrary values, remove restriction for values to be >= 0 and <= 1 * rename feature flag GRPC methods accordingly * distinguish between evaluating a probability feature flag and fetching the raw value
This also adds support for range feature flags in the feature flag service and its API.
[skip ci]
[skip ci]
it takes ages and is sometimes flaky and has zero value
Building the frontend container image on an Apple M1 would result in
the following error during docker build:
15.56 > Build error occurred
15.56 [Error: ENOENT: no such file or directory, copyfile '/app/node_modules/sharp/build/Release/sharp-darwin-arm64v8.node' -> '/app/.next/standalone/node_modules/sharp/build/Release/sharp-darwin-arm64v8.node'] {
15.56 errno: -2,
15.56 code: 'ENOENT',
15.56 syscall: 'copyfile',
15.56 path: '/app/node_modules/sharp/build/Release/sharp-darwin-arm64v8.node',
15.56 dest: '/app/.next/standalone/node_modules/sharp/build/Release/sharp-darwin-arm64v8.node'
15.56 }
Updating sharp to v0.33.x avoids this problem, as that version includes
pre-built sharp binaries for various platforms, see
https://sharp.pixelplumbing.com/changelog#v0330---29th-november-2023
[skip ci]
This effectively reverts aefb610. Having no arm64 images makes running the demo locally via K8s much harder, because K8s tries to pull the arm64 version of the images and that of course fails when it doesn't exist.
This allows to attach a persistent volume to the PostgreSQL service and still always run the init scripts on startup. For the first startup with a fresh volume, the database will be initialized correctly; on subsequent starts the init scripts will do nothing.
Either deploy to one namespace or to two different namespaces (just to make the deployment topology a bit more "interesting").
When introducing error rates/probabilities with 430b4c9, the correct move would have been to remove the 1/10 hard coded error probability in adservice (similar to b55b147 for cart service), since asking the feature flag service whether adServiceFailure is enabled is a random experiment each time anyway. Leaving the 1/10 probability in place in ad service as well skews the error rate by a factor of 1/10.
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
This reverts commit bbe3d38.
This functionality is already covered by ide-gen-proto.sh. [skip ci]
- automatically refresh image pull secret before trying to deploy - use both $dash0_configuration_dir/demo/values.yaml as well as $dash0_configuration_dir/demo/environments/aws/demo-eu-west-1-demo.yaml to produce the config for local deployment [skip ci]
Use a different cloud account id to be able to differentiate data from a local deployment from the actual demo deployment.
Extract port forward into separate script.
Do not report local data to prod, report only to dash0-dev.com.
The resource detectors are configured in a file loaded via a --require hook (in utils/telemetry/Instrumentation.js), which needs to be pure JS.
* remove deployment to two namespaces, we can resurrect that if we pick up https://linear.app/dash0/issue/ENG-951/make-the-deployment-topology-more-interesting again * rename namespace from otel-demo-ns to otel-demo * extract processing of values file into separate script * add support for sending data to a configurable ingress URL via OTEL_EXPORTER_OTLP_ENDPOINT * add support for using a custom values.yaml file via VALUES_YAML
* remove Grafana, Jaeger, Prometheus and OpenSearch by default
* add option to send data to recorder
* add recorder collector to write telemetry to files.
Too many random people (not Dash0 employees) requesting access. [skip ci]
github-actions
bot
added
the
helm-update-required
marcelbirkner
deleted the
eng-4410-fill-otel-demo-with-more-data-required-before-kubecon
branch
|
Will create a dedicated upstream PR. Needs more cleanup |
marcelbirkner
restored the
eng-4410-fill-otel-demo-with-more-data-required-before-kubecon
branch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.