Skip to content

ci: disable ccache and toolkit cache for self-hosted OpenVINO runners#20594

Closed
Copilot wants to merge 3 commits intomasterfrom
copilot/sub-pr-20593
Closed

ci: disable ccache and toolkit cache for self-hosted OpenVINO runners#20594
Copilot wants to merge 3 commits intomasterfrom
copilot/sub-pr-20593

Conversation

Copy link
Contributor

Copilot AI commented Mar 15, 2026

GitHub Actions cache causes issues on self-hosted runners due to cross-OS detection — runners alternately create their own cache entries, leading to conflicts. Both the ccache and OpenVINO toolkit cache need to be fully disabled for self-hosted runners.

  • build.yml: Added if: matrix.variant != 'gpu' to the ccache and Use OpenVINO Toolkit Cache steps. The gpu variant runs on self-hosted while cpu runs on ubuntu-24.04. The existing Setup OpenVINO Toolkit condition (cache-hit != 'true') naturally evaluates to true when the cache step is skipped, so the toolkit is always downloaded fresh on self-hosted.
  • build-self-hosted.yml: Removed the Use OpenVINO Toolkit Cache step entirely from ggml-ci-intel-openvino-gpu-low-perf (always self-hosted) and dropped the cache-hit condition on Setup OpenVINO Toolkit so it runs unconditionally.

💬 Send tasks to Copilot coding agent from Slack and Teams to turn conversations into code. Copilot posts an update in your thread when it's finished.

@CISC
Copy link
Member

CISC commented Mar 15, 2026

Well, that was fruitful...

Co-authored-by: ggerganov <1991296+ggerganov@users.noreply.github.com>
Copilot AI changed the title [WIP] [#15307] Fix ccache saving in OpenVINO on master branch ci: disable ccache and toolkit cache for self-hosted OpenVINO runners Mar 15, 2026
Copilot AI requested a review from ggerganov March 15, 2026 17:10
@CISC
Copy link
Member

CISC commented Mar 15, 2026

Ah, just a massive delay?

@ggerganov
Copy link
Member

Yup, very slow. copilot ngmi

But I wonder if the suggestion to remove the toolkit cache from the self-hosted workflows makes sense.

@CISC
Copy link
Member

CISC commented Mar 15, 2026

But I wonder if the suggestion to remove the toolkit cache from the self-hosted workflows makes sense.

Yep, will do once CI finishes.

@ggerganov
Copy link
Member

Even though the workflow is self-hosted, it seems the cache still works:

https://github.com/ggml-org/llama.cpp/actions/runs/23111057155/job/67128454815

So maybe we want to keep the toolkit cache active.

Not 100% sure.

@CISC
Copy link
Member

CISC commented Mar 15, 2026

Even though the workflow is self-hosted, it seems the cache still works:

It works if it was created by that os previously, otherwise it will create a new one, might as well disable it.

@ggerganov
Copy link
Member

Yes, lets disable. It's better to put a bit more load on that runner compared to taking a bit from the GH cache.

@CISC
Copy link
Member

CISC commented Mar 15, 2026

Even though the workflow is self-hosted, it seems the cache still works:

It works if it was created by that os previously, otherwise it will create a new one, might as well disable it.

The cache is named with runner.os, but that's not sufficient as actions/cache has further checks it seems.

Base automatically changed from cisc/ci-openvino-ccache-master to master March 15, 2026 17:58
@CISC CISC closed this Mar 15, 2026
@CISC CISC deleted the copilot/sub-pr-20593 branch March 15, 2026 18:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants