feat: improve s3 upload speed

[antoinetran]

Summary

What change needs making?

The S3 upload code used in artifact upload.

Use Cases

When would you use this?

In environment where workflows with large artifacts (>10GB) where stage out is a bit slow due to non optimal upload thread number/part size.

In such environment, one can configure with env variables inside the Argo Workflows ConfigMap the number of threads / part size that works well with a given S3 provider.

Benchmark and more detail in #15260

Message from the maintainers:

Love this feature request? Give it a 👍. We prioritise the proposals with the most 👍.

[coderabbitai]

🔗 Related PRs

#15124 - chore(ci): parallelize CLI builds in release workflow [merged]
#15265 - fix: support large container args [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!