Skip to content

Fix CloudRift API compatibility - #4380

Open
peterschmidt85 wants to merge 1 commit into
masterfrom
issue_4379_cloudrift_api
Open

peterschmidt85 wants to merge 1 commit into
masterfrom
issue_4379_cloudrift_api

Conversation

@peterschmidt85

Copy link
Copy Markdown
Contributor

Fixed #4379

@apedintensor

Copy link
Copy Markdown

Thanks for the quick fix! I tested this branch against the live CloudRift API with a local server (cloudrift backend only, fresh server dir):

  • dstack offer --gpu RTX5090 lists CloudRift offers as before.
  • A nodes: 0..1 fleet + a task on rtx59-7-24-800-ig.1 in ap_northeast_kr_se_1 ($0.65/h) provisioned the VM, ran nvidia-smi (RTX 5090, 32607 MiB, driver 580.159.03), and exited 0. Total about 11 minutes from submit, mostly VM boot.
  • With idle_duration: 0s, the instance was released afterwards. CloudRift's instances/list showed no active instances about 15 s after the run finished.

No 400 errors from recipes/list, instances/rent or instances/list.

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.

[Bug]: CloudRift provisioning fails with 400 because the API rejects pinned version 2025-05-29

2 participants