-
Notifications
You must be signed in to change notification settings - Fork 512
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[spark-rapids] generate spark-rapids/spark-rapids.sh from template #1284
base: master
Are you sure you want to change the base?
[spark-rapids] generate spark-rapids/spark-rapids.sh from template #1284
Conversation
/gcbrun |
/gcbrun |
/gcbrun |
…re cases ; defined utility function to verify installation via pyspark
/gcbrun |
/gcbrun |
…e non-default variants
/gcbrun |
/gcbrun |
/gcbrun |
/gcbrun |
/gcbrun |
/gcbrun |
okay so now we have some rocky8 and rocky9 coverage. I don't think we have to disable the 2.0 images. Let's verify... |
* defining is_debuntu with the other os comparison functions * refactored gpu-related code out of common function library * being more surgical about signing material usage * skip installing dependencies if it has already been done * refactored configure_gpu_exclusive_mode to fewer lines * not relying on dkms certs being deployed to use modulus_md5sum * removed dependency on pciutils * less reactive to not having GPU
/gcbrun |
…nd so installation should succeed
okay, now it:
Let's try skipping fewer tests still |
/gcbrun |
… version is fine where it is at
/gcbrun |
* 40G is plenty ; no need for 50G
I think we've got single node clusters working fine on rocky again. |
/gcbrun |
can we skip no tests? |
/gcbrun |
/gcbrun |
/gcbrun |
/gcbrun |
Failing on 2.1-rocky8:
|
The one on 2.0-rocky8 seems slightly different:
|
/gcbrun |
This is a re-implementation of the script using templates created while re-factoring common code under gpu/, dask/, rapids/, spark-rapids/, horovod/, mlvm/ and many of the other initialization actions.