Collect project metrics through Prometheus client

[nikimanoledaki]

This issue details what needs to be done in the collect step.

You can reference the proposal 3 - Metrics section: https://github.com/cncf-tags/green-reviews-tooling/blob/main/docs/proposals/proposal-003-report.md#metrics.

There are two parts to this issue: setting up a Go client, and fetching project metrics from Prometheus.

1. Setup the Go client that fetches queries from the Prometheus HTTP API

Create a small Go client that runs in the cluster and authenticates with Prometheus to fetch metrics through queries.

Acceptance criteria:

• Connect to the local Prometheus API via its address - example
• Perform a query

2. Collect project metrics

The metrics we will need to collect through the Prometheus client are:

container_cpu_usage_seconds_total
container_memory_rss
container_memory_working_set_bytes

Note: Since we don't have a storage solution yet, it's okay to simply collect the metrics and log the outcome for now.

3. Fetch metrics as part of benchmark-pipeline.yaml

Acceptance criteria:

• The Prometheus client should fetch these metrics at the end of a benchmark-job.
• The timestamp of the Prometheus query should be the same as the start and end of the benchmark job.

[nikimanoledaki]

@dipankardas011 please leave a comment so that I can assign you to it, thank you!!

[dipankardas011]

yes I can work on this

[nikimanoledaki]

@dipankardas011, awesome, it's all yours! Feel free to create multiple PRs etc - it'll be easier to review that way :) I added more information to the description.