Skip to content

Observability enhancement for HiCache - #32388

Merged
xiezhq-hermann merged 3 commits into
mainfrom
hicache-l2-telemetry
Aug 5, 2026
Merged

xiezhq-hermann merged 3 commits into
mainfrom
hicache-l2-telemetry

Conversation

@xiezhq-hermann

@xiezhq-hermann xiezhq-hermann commented Jul 25, 2026

Copy link
Copy Markdown
Collaborator

Motivation

This PR introduce several new metrics for HiCache observability including: separated cache hit counter from tiers, backuped token, load back token and dropped token (for write-back policy) counter, more accurate timing and memory footprint counter for live bandwidth computation.

Modifications

Accuracy Tests

Speed Tests and Profiling

Checklist

Review and Merge Process

  1. Ping Merge Oncalls to start the process. See the PR Merge Process.
  2. Get approvals from CODEOWNERS and other reviewers.
  3. Trigger CI tests with comments or contact authorized users to do so.
    • Common commands include /tag-and-rerun-ci, /tag-run-ci-label, /rerun-failed-ci
  4. After green CI and required approvals, ask Merge Oncalls or people with Write permission to merge the PR.

CI States

Latest PR Test (Base): ✅ Run #30943224192
Latest PR Test (Extra): ✅ Run #30965794215

@gemini-code-assist

Copy link
Copy Markdown
Contributor

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

@xiezhq-hermann
xiezhq-hermann requested a review from bingxche as a code owner July 27, 2026 10:10
@github-actions github-actions Bot added documentation Improvements or additions to documentation amd dependencies Pull requests that update a dependency file lora Multi-modal multi-modal language model deepseek npu diffusion SGLang Diffusion jit-kernel labels Jul 27, 2026
@xiezhq-hermann xiezhq-hermann added run-ci bypass-fastfail and removed documentation Improvements or additions to documentation amd dependencies Pull requests that update a dependency file Multi-modal multi-modal language model deepseek hicache Hierarchical Caching for SGLang npu diffusion SGLang Diffusion jit-kernel lora labels Jul 27, 2026
@github-actions github-actions Bot added the hicache Hierarchical Caching for SGLang label Jul 27, 2026
xiezhq-hermann and others added 3 commits August 3, 2026 17:28
Add metrics for the HiCache L2 (host) tier covering both HiRadixCache
and UnifiedRadixCache:

- Per-pool token counters and byte counters for D->H backup and
  H->D load-back, with sizes recorded for bandwidth calculation
- Duration histograms for backup and load-back operations
- Dropped-token counter for KV subtrees discarded without backup
  under host memory pressure
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants