Add blog post on AKS Configurable Scheduler Profiles

[colinmixonn]

This blog post introduces AKS Configurable Scheduler Profiles, highlighting their benefits for optimizing resource utilization and improving scheduling strategies for web-distributed and AI workloads. It covers configuration examples for GPU and CPU utilization.

[Copilot]

Pull request overview

This pull request adds a new blog post announcing the preview of AKS Configurable Scheduler Profiles, a feature that enables fine-grained control over pod scheduling strategies to optimize resource utilization and improve workload performance.

Key Changes

• Introduces a new "scheduler" tag to categorize blog posts related to pod placement and scheduling optimization
• Adds comprehensive blog post covering three main scheduling use cases: GPU bin-packing for AI workloads, pod distribution across topology domains for resilience, and memory-optimized scheduling with PVC-aware placement
• Provides YAML configuration examples and best practices for implementing custom scheduler profiles

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 20 comments.

File	Description
website/blog/tags.yml	Adds new "scheduler" tag for categorizing posts about pod placement and scheduling techniques
website/blog/2025-12-16-aks-config-scheduler-profiles-preview/index.md	New blog post introducing AKS Configurable Scheduler Profiles with configuration examples for GPU utilization, topology distribution, and memory-optimized scheduling

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 9 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 4 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 5 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 7 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 5 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 1 comment.

[kaarthis]

Good write up but opportunities identified to make it better .

[kaarthis]

What year are these CastAI and Datadog reports from? Cite the year. Stale stats kill credibility with senior platform engineers. Also — the blog headlines GPU utilization but the hook is CPU-only. Add one GPU underutilization stat (NVIDIA publishes these, or use internal AKS telemetry) so the AI/ML reader is hooked from line 1.

[kaarthis]

Good foundational content. But — does the external customer reading a blog about configurable profiles need 20 lines on how the default scheduler works? Consider trimming to 5-6 lines + diagram + a link to upstream docs for "deep dive on scheduling cycles." Get to the value prop faster.

[colinmixonn]

This section is a part of the value prop because the reader first needs to understand how the default scheduler works to understand the config scheduler.

The value prop is raised in the intro and with the TOC links if it's already known the reader can skip.

[kaarthis]

The YAML disables PodTopologySpread:
Why is this not explained? This is a significant decision. A customer who applies this profile loses zone-aware spreading. If they have topology spread constraints in their pod spec, those are silently ignored by this scheduler profile. This needs a prominent warning:

"We disable PodTopologySpread in this profile because bin-packing and zone-spreading are opposing goals. If you need both high utilization AND zone resilience, use separate named profiles for different workload classes (e.g., cpu-binpacking-scheduler for batch, default-scheduler for HA services)." This directly connects to Wilson's NAP blog which covers topology spread in depth. Cross-reference opportunity.

[kaarthis]

same concern as before why disabled and explain . Even more critical for GPU workloads — if a customer has multi-zone GPU nodes and disables topology spread, all inference pods could land in one zone. One zone outage = total inference downtime. This needs a warning.

[kaarthis]

FAQ #2 repeats Wilson's content less thoroughly. The customer who reads both blogs will find redundancy.
Is it a blog consolidation opportunity - pls check with wilson and Ahmed.

[colinmixonn]

Wilsons blog will only Focus on NAP. This section is to provide some answers to the immediate question of using any autoscaler with config scheduler. Wilson and i are in conversation about cross-referencing

[kaarthis]

Missing Section: Upstream Kubernetes Work & AKS Roadmap - you can point to KEPs and work being done and planned too.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 8 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 6 changed files in this pull request and generated 2 comments.

[Copilot]

Typo in the table of contents link text: "ncrease" should be "increase".

Suggested change

	2. [Use Configurable Scheduler Profiles to ncrease node utilization and operator control](#configurable-scheduler-profiles-on-aks)
	2. [Use Configurable Scheduler Profiles to increase node utilization and operator control](#configurable-scheduler-profiles-on-aks)

[Copilot]

Missing space after the period before the "Configure node bin-packing" link, which will read as a single word/sentence fragment when rendered.

Suggested change

	When combined, these plugins favor GPU‑bound nodes with balanced CPU and memory usage over nodes with large amounts of unused memory or fragmented resources. This results in more efficient GPU placement and fewer partially utilized nodes.[Configure node bin-packing][configure-most-allocated] using the MostAllocated strategy to improve utilization and reduce infrastructure costs.
	When combined, these plugins favor GPU‑bound nodes with balanced CPU and memory usage over nodes with large amounts of unused memory or fragmented resources. This results in more efficient GPU placement and fewer partially utilized nodes. [Configure node bin-packing][configure-most-allocated] using the MostAllocated strategy to improve utilization and reduce infrastructure costs.