AndreaBozzo commented on PR #487:
URL: https://github.com/apache/hudi-rs/pull/487#issuecomment-3679285037

   Hi @xushiyan, thanks for the feedback! You're right to question the gains
   
   We hit several issues (permission errors, missing mc image) that required 
multiple fix iterations. The CI runs you saw were failing/suboptimal.
   
   **Intended gains (once working):**
   - Build Python bindings once → reuse across 4 integration tests (currently 
each rebuilds from scratch)
   - ~6-9 min savings on code PRs (3 redundant `make setup develop` eliminated)
   
   **Current overhead concerns:**
   - Artifact upload/download (~1-2GB runner image)
   - Still need to build `mc` in each job
   
   **Options:**
   1. **Simplify** - Drop the artifact approach, just use GitHub Actions cache 
for cargo/target dirs
   2. **Abandon** - Keep main branch approach if complexity > benefit
   
   I'm perfectly fine with both, as the change was meant to be beneficial, not 
detrimental of course. 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to