Skip to content
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

[HOLD][npcollector] Fix statsd client #25687

Closed
wants to merge 1 commit into from

Conversation

AlexandreYang
Copy link
Member

@AlexandreYang AlexandreYang commented May 17, 2024

Superseded by #25708

What does this PR do?

[npcollector] Fix statsd client

statsd.Client is assigned/configured at runtime, currently we might get the NoOpClient instead of the actual statsdClient. This PR fixes that issue by wrapping inside a function.

// Client is a global Statsd client. When a client is configured via Configure,
// that becomes the new global Statsd client in the package.
var Client statsd.ClientInterface = &statsd.NoOpClient{}
// Configure creates a statsd client from a dogweb.ini style config file and set it to the global Statsd.
func Configure(host string, port int, create func(string, int, ...statsd.Option) (statsd.ClientInterface, error)) error {
var err error
Client, err = create(host, port)
return err
}

Motivation

Additional Notes

Possible Drawbacks / Trade-offs

Describe how to test/QA your changes

Base automatically changed from NETPATH-73-np-scheduler to main May 17, 2024 08:55
@AlexandreYang AlexandreYang marked this pull request as ready for review May 17, 2024 09:01
@AlexandreYang AlexandreYang requested review from a team as code owners May 17, 2024 09:01
@pr-commenter
Copy link

pr-commenter bot commented May 17, 2024

Test changes on VM

Use this command from test-infra-definitions to manually test this PR changes on a VM:

inv create-vm --pipeline-id=34535645 --os-family=ubuntu

@pr-commenter
Copy link

pr-commenter bot commented May 17, 2024

Regression Detector

Regression Detector Results

Run ID: 3ff701ad-2245-4bbc-8c22-fe698c8b38a2
Baseline: 0e8fe47
Comparison: ebd2611

Performance changes are noted in the perf column of each table:

  • ✅ = significantly better comparison variant performance
  • ❌ = significantly worse comparison variant performance
  • ➖ = no significant change in performance

No significant changes in experiment optimization goals

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf experiment goal Δ mean % Δ mean % CI
basic_py_check % cpu utilization +2.00 [-0.55, +4.55]
uds_dogstatsd_to_api_cpu % cpu utilization +0.97 [-1.90, +3.84]
idle memory utilization +0.34 [+0.30, +0.39]
file_tree memory utilization +0.08 [-0.04, +0.20]
otel_to_otel_logs ingress throughput +0.04 [-0.32, +0.39]
uds_dogstatsd_to_api ingress throughput +0.01 [-0.19, +0.21]
trace_agent_json ingress throughput +0.00 [-0.01, +0.01]
trace_agent_msgpack ingress throughput -0.00 [-0.00, +0.00]
tcp_dd_logs_filter_exclude ingress throughput -0.03 [-0.07, +0.01]
tcp_syslog_to_blackhole ingress throughput -0.71 [-21.80, +20.38]
pycheck_1000_100byte_tags % cpu utilization -2.91 [-7.59, +1.78]

Explanation

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

  1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

  2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

  3. Its configuration does not mark it "erratic".

@hmahmood
Copy link
Contributor

I am not sure what problem this PR is solving. When are you seeing that you get the no-op client?

@@ -76,6 +76,9 @@ func newNpCollectorImpl(epForwarder eventplatform.Forwarder, collectorConfigs *c
collectorConfigs.pathtestInterval,
collectorConfigs.flushInterval)

getStatsdClient := func() ddgostatsd.ClientInterface {
return statsd.Client
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@hmahmood (starting a thread since it's easier to follow discussions)

I am not sure what problem this PR is solving. When are you seeing that you get the no-op client?

When I log the statsd.Client it displays this:

2024-05-17 08:24:04 PDT | PROCESS | INFO | (comp/networkpath/npcollector/npcollectorimpl/npcollector.go:79 in newNpCollectorImpl) | statsd.Client &statsd.NoOpClient{}

I think it's because the NpCollector component is created before statsd Configure() is called:

func Configure(host string, port int, create func(string, int, ...statsd.Option) (statsd.ClientInterface, error)) error {
var err error
Client, err = create(host, port)
return err
}

Also, I'm not seeing any statsd metrics being reported due to use of NoOpClient.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I haven't worked much with fx -- should we be injecting the statsd client? In general DI frameworks are supposed to solve these configuration ordering issues.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi @leeavital , I think you are right, statsd client should be injected as component instead of using a global (currently).

I will reach processes team to see if that's something we can do.

@AlexandreYang AlexandreYang marked this pull request as draft May 20, 2024 14:24
@AlexandreYang AlexandreYang changed the title [npcollector] Fix statsd client [HOLD][npcollector] Fix statsd client May 20, 2024
@AlexandreYang
Copy link
Member Author

Superseded by #25708

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants