I propose two improvement for the data format.
Proposal1:
Some users don't need (or want) to specify “targetNode” every time
- It is tedious to specify for each node when there are many servers.
- User only needs to have the number of GPUs in the entire K8s cluster(number of GPUs per node is not important for this user)
To satisfy this demand, we might need to design something like this:
- The user specifies the number of GPUs in the cluster itself.
- After that, the CDI Operator may need to actually determine the number of GPUs for each node
Of course, we will keep the existing functionality so that users can still directly specify the number of GPUs per node.
Proposal2:
Allows user to specify a variety of device types.
GPU, PCIeSSD, NIC, CXL Memory, etc.
I propose two improvement for the data format.
Proposal1:
Some users don't need (or want) to specify “targetNode” every time
To satisfy this demand, we might need to design something like this:
Of course, we will keep the existing functionality so that users can still directly specify the number of GPUs per node.
Proposal2:
Allows user to specify a variety of device types.
GPU, PCIeSSD, NIC, CXL Memory, etc.