DCGM icon indicating copy to clipboard operation
DCGM copied to clipboard

how can I clear stale XID error

Open zdyang opened this issue 1 year ago • 0 comments

We employ the dcgm-exporter to monitor our GPU cluster. Occasionally, we come across an XID error referred to as "XID 31". This error is typically caused by a user program. Interestingly, even after exiting the program, the XID error persists. I am curious if there is a method to resolve this outdated XID error. Thank you for your assistance.

zdyang avatar Oct 07 '23 07:10 zdyang