The start StartSpotPod endpoint is missing the gpuCount argument, causing it to not work at all. I added the argument to the ctl, and the corresponding graphQL call.
How I tested it
I run the graphQL query by hand with and without the argument I added. Without it, I get back a 400 error, and with it everything works great.
Add gpuCount argument
Related:
156: can't start spot pod using the cli
Description
The start StartSpotPod endpoint is missing the gpuCount argument, causing it to not work at all. I added the argument to the ctl, and the corresponding graphQL call.
How I tested it
I run the graphQL query by hand with and without the argument I added. Without it, I get back a 400 error, and with it everything works great.