mirror of
https://github.com/kvcache-ai/Mooncake.git
synced 2026-11-05 01:17:55 +00:00
[GH-ISSUE #1266] [Usage]: How to use the provided trace? #422
Labels
No labels
ascend
auto-closed
bug
bug
CI/Build
Common
documentation
enhancement
feature request
good first issue
help wanted
high priority
MetaX
nightly-failure
pull-request
PyTorch Backend
question
RFC
Roadmap
stale
Store
Tests
Transfer Engine
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
kvcache-ai/Mooncake#422
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @GDLTZzq on GitHub (Dec 23, 2025).
Original GitHub issue: https://github.com/kvcache-ai/Mooncake/issues/1266
Describe your usage question
Trace provides different stored hash_id, but there is no relevant actual input. How to use the provided trace? Can you provide an example benchmark?
Before submitting a new issue...
@ykwd commented on GitHub (Dec 24, 2025):
Due to privacy and security concerns, we are not able to release the original traces.
From my personal experience, when running experiments with traces, you can roughly approximate the original trace by randomly generating the same inputs for the same hash ID.
@GDLTZzq commented on GitHub (Dec 24, 2025):
I see. I also want to ask, will the tested code be released in the future, that is, the code to realize the above logic?
@ykwd commented on GitHub (Dec 24, 2025):
I didn’t keep the code, but the implementation logic is very simple, and it should be easy to generate one quickly using AI. Of course, I’m mainly speaking from the perspective of system performance testing. The randomly generated inputs produced using the above logic can be used to simulate request lengths and KV-cache reuse patterns.
@GDLTZzq commented on GitHub (Dec 24, 2025):
OK,I got it ,Thank you for your reply