[Bugfix] Fix incorrect kv cache metrics in grafana.json #27133
Conversation
|
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run You ask your reviewers to trigger select CI tests on top of Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add If you have any questions, please reach out to us on Slack at https://slack.vllm.ai. 🚀 |
|
Documentation preview: https://vllm--27133.org.readthedocs.build/en/27133/ |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates the grafana.json file to use the correct Prometheus metric name for GPU cache usage. The old metric name gpu_cache_usage_perc is replaced with the new metric name kv_cache_usage_perc. This change ensures that the Grafana dashboard accurately reflects the current metrics being exposed by vLLM.
| "uid": "${DS_PROMETHEUS}" | ||
| }, | ||
| "editorMode": "code", | ||
| "expr": "vllm:gpu_cache_usage_perc{model_name=\"$model_name\"}", |
There was a problem hiding this comment.
The metric name vllm:gpu_cache_usage_perc is outdated and should be updated to vllm:kv_cache_usage_perc to align with the latest metrics. This discrepancy could lead to incorrect monitoring and alerting.
It's critical to ensure that monitoring dashboards use the correct metric names to provide accurate insights into system performance and resource utilization.
"expr": "vllm:kv_cache_usage_perc{model_name=\"$model_name\"}"
|
Thank you! Just need to fix the DCO issue:
Since you'll be making that update, would you mind also searching and replacing in |
Signed-off-by: Fangping Shi <[email protected]>
fangpings
force-pushed
the
fix_grafana
branch
from
d98eae4 to
f3c5ca8
Compare
Signed-off-by: Fangping Shi <[email protected]>
Thanks. I replaced occurrence under that folder too. |
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]>
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]> Signed-off-by: Alberto Perdomo <[email protected]>
…o step_forward * 'step_forward' of https://github.com/raindaywhu/vllm: (148 commits) [Model] Add MoE support for NemotronH (vllm-project#25863) [Metrics] [KVConnector] Add connector prefix cache hit rate stats (vllm-project#26245) [CI] Reorganize entrypoints tests (vllm-project#27403) add SLA information into comparison graph for vLLM Benchmark Suite (vllm-project#25525) [CI/Build] Fix AMD CI: test_cpu_gpu.py (vllm-project#27388) [Bugfix] Fix args settings for guided decoding args (vllm-project#27375) [CI/Build] Fix Prithvi plugin test (vllm-project#27393) [Chore] Remove duplicate `has_` functions in vllm.utils (vllm-project#27372) [Model] Add num_cached_tokens for PoolingRequestOutput (vllm-project#27378) [V1][spec decode] return logprobs for spec decoding (vllm-project#26060) [CORE] Support Prefix Caching with Prompt Embeds (vllm-project#27219) [Bugfix][Core] running queue index leakage exception (vllm-project#26754) [Bugfix] Fix incorrect kv cache metrics in grafana.json (vllm-project#27133) [Bugfix] Fix SLA tuner initialization (vllm-project#27355) [Bugfix] Fix deepseek-ocr multi-image inference and add `merge_by_field_config=True` with tensor schema support (vllm-project#27361) [MLA] Bump FlashMLA (vllm-project#27354) [Chore] Separate out system utilities from vllm.utils (vllm-project#27201) [BugFix] bugfix for Flash Attention MLA with full cuda graph IMA following pr-25490 (vllm-project#27128) [Feature] publisher default set zmq in kv_event config (vllm-project#26915) [Prefix Cache] Use LoRA name for consistent KV-cache block hashing (vllm-project#27211) ...
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]>
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]> Signed-off-by: 0xrushi <[email protected]>
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]> Signed-off-by: 0xrushi <[email protected]>
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]>
…#27133) Signed-off-by: Fangping Shi <[email protected]> Co-authored-by: Fangping Shi <[email protected]>
4 participants
Purpose
Current grafana.json is using an outdated prometheus metrics name
gpu_cache_usage_perc. The lastest metrics is namedkv_cache_usage_perc. See hereTest Plan
Verified in my own grafana dashboard
Test Result
Before
After
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.