Startup script runs in parallel to the deep learning machine image startup

108 Views Asked by At

I am using the deep learning image, and also submit a startup script at the same time to take care of some customization but it appears to run in parallel which causes some lock conditions. Is there a recommended work around?

1

There are 1 best solutions below

0
On

In the release m20 (has been released on Feb 22 2019) this problem has been fixed. There is new metadata value: post-startup-script any HTTPS or GS url can be set as value for the variable and it will be executed after all the initial scripts are finished.