Felix Abecassis

You are both correct :) @alexellis > My question was: "could you run two different functions using [the same] GPU at the same time?" (expecting an answer of no) and...

I don't think you should try to hack around the official upstream support: that means don't overcommit GPUs. If you need to run multiple pods for the same function, you...

I think you should embrace the current upstream support, including its limitations. If you assume that the cluster is already configured with the NVIDIA device driver, the device plugin and...

@sberryman yes, our device plugin implementation supports multiple GPUs on one node and set this environment variable accordingly for the container.

Yes I know it's the default, I explicitly set `--invalidate=1` in the example above to show that it's the problem and in case `fio` changes/changed the default behavior across releases....

The fio debug log doesn't seem very helpful, particularly since it doesn't include timestamps: ``` file 224907 dup files: 0 io 224907 load ioengine psync file 224907 add file socket0.0.0...

https://github.com/opencontainers/runtime-spec/pull/673#discussion_r99942858 > then the runtime errors out. end-users investigate. complain to the runtime developers or write a fix. That kind-of linting and validation needs to be secondary to this spec....

> The specification is basically "the input format for libseccomp" so if you want to specify the behaviour differently you might consider first writing a new seccomp library for runc,...

> By the way, was there a reason to not have lxc.mount.entry immediately after lxc.mount.fstab? Should we try to fix it or is too late and part of the implicit...

Was out of the office for a few days, I'll try to do it this week.