hard20 pts · first answer countsmodel-servingdeploymentkubernetes
Rolling update stuck in Pending
A Kubernetes rolling update of a GPU deployment with maxSurge=1 stalls: the new pod is Pending, and all GPUs are in use by the old pods. What is the best fix?
Discussion 0
Join free to post in the discussion.
Related questions
- When batch inference fitsMLOps Engineer
- Canary deploymentMLOps Engineer
- Traffic before the model has loadedMLOps Engineer
- Why export a model to ONNXMLOps Engineer
- Protocol for internal inference callsMLOps Engineer
- Pod stuck in CrashLoopBackOffMLOps Engineer
- Asking Kubernetes for a GPUMLOps Engineer
- GPU nodes show zero GPUsMLOps Engineer