-
Notifications
You must be signed in to change notification settings - Fork 1.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CONTINT-53] Make ECS e2e tests target the fake intake #20286
Conversation
f82d054
to
e48ef5e
Compare
e48ef5e
to
20b3c67
Compare
Bloop Bleep... Dogbot HereRegression Detector ResultsRun ID: 7d3441f7-41f2-49e6-81cf-c81df04c84d6 ExplanationA regression test is an integrated performance test for Because a target's optimization goal performance in each experiment will vary somewhat each time it is run, we can only estimate mean differences in optimization goal relative to the baseline target. We express these differences as a percentage change relative to the baseline target, denoted "Δ mean %". These estimates are made to a precision that balances accuracy and cost control. We represent this precision as a 90.00% confidence interval denoted "Δ mean % CI": there is a 90.00% chance that the true value of "Δ mean %" is in that interval. We decide whether a change in performance is a "regression" -- a change worth investigating further -- if both of the following two criteria are true:
The table below, if present, lists those experiments that have experienced a statistically significant change in mean optimization goal performance between baseline and comparison SHAs with 90.00% confidence OR have been detected as newly erratic. Negative values of "Δ mean %" mean that baseline is faster, whereas positive values of "Δ mean %" mean that comparison is faster. Results that do not exhibit more than a ±5.00% change in their mean optimization goal are discarded. An experiment is erratic if its coefficient of variation is greater than 0.1. The abbreviated table will be omitted if no interesting change is observed. No interesting changes in experiment optimization goals with confidence ≥ 90.00% and |Δ mean %| ≥ 5.00%. Fine details of change detection per experiment.
|
20b3c67
to
2f881d3
Compare
if err != nil { | ||
collect.Errorf("%w", err) | ||
return | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
💭 thought
Looking forwards to stretchr/testify#1481
func (suite *ecsSuite) TestNginx() { | ||
// `nginx` check is configured via docker labels | ||
// Test it is properly scheduled | ||
suite.testMetric("nginx.net.request_per_s", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
❓ question
Have you considered having an helper function instead of using a suite
as a base ? Something like
func assertMetrics(t *testing.T, fakeintake *fakeintake.Client, metrics MetricsToCheck)
Not a critic, I am curious if you considered it as an alternative, as here we use the suite to pass the testing and fakeintake context rather than to leverage the suite interface.
An alternative could also be to pass a testing.T
context to fakeintake.Client
. But I can see this has the limit of not allowing switching testing context when calling it from an Eventually
or from a t.Run
that create a separate testing context.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
To be honest, I’m not convinced by the current way the helper functions are designed. And I know I’ll refactor them at some point.
For ex., the testMetrics
function takes a list of expected tags in the form of regular expressions. I think the code might be less painful to write if the calls to regexp.MustCompile(…)
were done inside the testMetrics(…)
function itself rather than on the caller side, everywhere this function is invoked.
In the current tests, testMetrics(…)
is only checking for the tags on the metrics.
Its parameters are:
- the name of the metric to look at
- a list of tags to filter the series to consider
- the exhaustive list of tags we expect to have on the series.
At some point, we’ll want to potentially check the value as well. As values are float, it will be an expected range.
It means two more parameters: - expected minimum value
- expected maximum value.
Having all that as positional arguments will become painful.
So, we might want to have named arguments, similar to the Pulumi API, with some arguments being optional:
assertMetric(&assertMetricArgs{
filter: {
metricName: …,
tags: …,
},
expect: {
tags: …,
value: {
min: …,
max: …,
})
About using a helper function instead of a baseSuite
method, I don’t have any strong opinion. It permitted to pass less parameters as the fakeintake object was a field of baseSuite
.
I think that I’d like to keep the fact that assertMetric
creates a sub-test with suite.Run(…)
because it permits to have a more granular report.
So, yeah, I agree that this needs improvement.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I also like how suite.Run
or t.Run
improve readability of reports, I do not use them for the separate context, still it can be misleading to use the parent context instead of the current one.
Thinking if we can have this in a generic FilterMetrics
inside the fakeintake, or if we can have some external aggregation helpers.
What ddev
does in integrations-core tests is having an aggregator that marks each checked metric and eventually can provide all metrics that were not checked with agg.RequireAllMetricsChecked
.
What does this PR do?
Move the
testMetric
function away from the Kubernetes-specifick8sSuite
object so that it can be used for non-Kubernetes tests.Motivation
Be able to use the same function to assert metrics for ECS and K8S scenarios.
Additional Notes
Possible Drawbacks / Trade-offs
Describe how to test/QA your changes
Reviewer's Checklist
Triage
milestone is set.major_change
label if your change either has a major impact on the code base, is impacting multiple teams or is changing important well-established internals of the Agent. This label will be use during QA to make sure each team pay extra attention to the changed behavior. For any customer facing change use a releasenote.changelog/no-changelog
label has been applied.qa/skip-qa
label is not applied.team/..
label has been applied, indicating the team(s) that should QA this change.need-change/operator
andneed-change/helm
labels have been applied.k8s/<min-version>
label, indicating the lowest Kubernetes version compatible with this feature.