feat: multi-format local sinks and S3 Parquet export for flows

[jpinsonneau]

Summary

• Default local flows output is json,sqlite; include parquet in --format only when needed (comma-separated / repeatable flags).
• Optional Hive-partitioned Parquet export to S3-compatible storage via embedded FLP encode.s3 (--s3-endpoint, --s3-bucket, --s3-account, credentials env/file/secret).
• Docs and help updated; e2e coverage for format/S3 YAML options.

Hub PR (operator): netobserv/netobserv-operator#2845

Custom images (required for S3 Parquet): There is no eBPF agent PR — only an FLP vendor bump. Until agent changes land separately, use the Quay tags below (agent embeds FLP with Parquet S3 encode; CLI image has this branch’s collector/format support):

Image	Tag
Agent	quay.io/jpinsonn/netobserv-ebpf-agent:s3-flowbuffer
CLI / collector	quay.io/jpinsonn/network-observability-cli:s3-flowbuffer

Override via env (see README / scripts/functions.sh):

export NETOBSERV_AGENT_IMAGE=quay.io/jpinsonn/netobserv-ebpf-agent:s3-flowbuffer
export NETOBSERV_COLLECTOR_IMAGE=quay.io/jpinsonn/network-observability-cli:s3-flowbuffer

(NETOBSERV_AGENT_IMAGE replaces the DaemonSet agent image; NETOBSERV_COLLECTOR_IMAGE is the in-cluster collector pod image.)

Test plan

• go test ./cmd/... ./e2e/...
• Local: flows --format=json,sqlite (default) and --format=parquet
• Kind + MinIO (below)

Kind testing — CLI

1. Kind cluster with MinIO reachable from agent pods (e.g. http://minio.minio.svc:9000).
2. Export custom images (required for S3 encode):

   export NETOBSERV_AGENT_IMAGE=quay.io/jpinsonn/netobserv-ebpf-agent:s3-flowbuffer
   export NETOBSERV_COLLECTOR_IMAGE=quay.io/jpinsonn/network-observability-cli:s3-flowbuffer

3. Credentials (do not pass secrets as CLI flags):

   export NETOBSERV_S3_ACCESS_KEY=...
   export NETOBSERV_S3_SECRET_KEY=...   # or AWS_* / credentials file / --s3-secret as documented

4. Capture:

   ./build/oc-netobserv flows \
     --s3-endpoint=http://minio.minio.svc:9000 \
     --s3-bucket=... \
     --s3-account=cli-test \
     --max-time=1m

5. Verify with MinIO client, e.g. mc ls under cluster_id=cli-test/ (Hive layout year=/month=/day=/hour=/part-….parquet); optional duckdb read_parquet('.../**/*.parquet').
6. Local defaults: --format=json,sqlite; include parquet in --format only when local Parquet parts are needed. S3 export is independent of local --format.

Notes

• S3 is wired on the agent via FLP_CONFIG (encode.s3), not collector --options. Missing credentials fail before deploy.
• Default S3 flush interval is 15s (capped by --max-time); agents flush pending Parquet parts on SIGTERM.

[openshift-ci]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign jotak for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• DOWNSTREAM_OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[openshift-ci]

PR needs rebase.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.