coldzerofear

Results 9 comments of coldzerofear

Fix device share plugins npe

/assign @k82cn

Fix enabling vgpu node lock

/assign @shinytang6

fix enable node device score plugin

/assign @shinytang6

Fix some potential null pointer panic

/assign @lowang-bh

Fix some potential null pointer panic

> > volcano.sh/volcano/pkg/scheduler/api/devices/nvidia/vgpu.(*GPUDevices).GetStatus(0xc0007173c0?) > > /go/src/volcano.sh/volcano/pkg/scheduler/api/devices/nvidia/vgpu/metrics.go:71 +0x18 > > The logs show that it panic here, we should also assert them. The assertions have been resolved on the outer layer,...

fix enable node device score plugin

/assign @thor-wl

[Bug]HAMi Framework Fails to Execute nanoGPT with CUDA, While NVIDIA k8s-device-plugin Succeeds

You can use the environment variable `LIBCUDA_LOG_LEVEL` to increase the logging level of the hami core and obtain more context

[Bug]HAMi Framework Fails to Execute nanoGPT with CUDA, While NVIDIA k8s-device-plugin Succeeds

> 在将`LIBCUDA_LOG_LEVEL``4` > > ``` > (base) (⎈|N/A:N/A)➜ cat output.txt | grep -i error > [HAMI-core Debug(492:140563747359616:hook.c:293)]: loading nvmlErrorString:2 > [HAMI-core Debug(492:140563747359616:hook.c:293)]: loading nvmlDeviceClearEccErrorCounts:10 > [HAMI-core Debug(492:140563747359616:hook.c:293)]: loading nvmlDeviceGetDetailedEccErrors:38 >...

1 node(s) had untolerated taint {node.kubernetes.io/unreachable: }. preemption: 0/4 nodes are available: 1 No preemption victims found for incoming pod, 3 Preemption is not helpful for scheduling..

If there are only two physical GPUs on your node, a single container can only request a maximum of 2 VGPUs