Skip to content

Conversation

@RasmusOrsoe
Copy link
Collaborator

@RasmusOrsoe RasmusOrsoe commented Jan 9, 2026

While #861 appeared to be solved in #860, updates to #857 showed that new variations in the workflow disk usage made the runner fail again.

These variations are largely out of our control; caused by infrastructure changes in GitHub and dependency changes.

In this PR, an additional step is added in the IceTray runner installation procedure, that avoid downloading cuda-versions of torch. This reduces the dependency-induced variation in disk usage, and the disk requirements by the installation from 6gb - 2gb.

This should create enough headroom for the workflow to be robust against future variations.

@RasmusOrsoe RasmusOrsoe changed the title 2nd workflow fix 2nd attempt at disk usage reduction for IceTray Runner Jan 9, 2026
Copy link
Collaborator

@Aske-Rosted Aske-Rosted left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

seems pretty non-controversial to me. Hopefully we do not have to think about runner disk space for a while now.

@RasmusOrsoe RasmusOrsoe merged commit f2a6c29 into graphnet-team:main Jan 11, 2026
12 checks passed
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.

2 participants