H2O Driverless AI Helm (GPU version) - unable to mount volume

[alanchia67]

Helm h2o-dai V 1.5.4
ICP 3.2.0 Enterprise Edition, ppc64le

At deployment, the below is observed in the events:

Unable to mount volumes for pod "h2o-gpu-h2o-dai-7ccbf6d757-zjf2w_h2o-dai(89738ff1-8e5c-11e9-8249-506b8df7ca6a)": timeout expired waiting for volumes to attach or mount for pod "h2o-dai"/"h2o-gpu-h2o-dai-7ccbf6d757-zjf2w". list of unmounted volumes=[config-vol]. list of unattached volumes=[logs-volume nvidia-pers-indicator config-vol default-token-tnl8m]

[jstzsyf]

I found a bug in templates/deployment.yaml of this chart:
Change "name: dai-filebeate-sidecar-config" to "dai-filebeat-sidecar-config-{{ .Release.Name }}".
I was able to deploy it on my ICP cluster after this change.

Btw, the h2o dai 1.6.2 was just released in docker hub. Can use "tag: 9.0-1.6.2" to get the latest image.

[brucesemple]

We removed the filebeat sidecar -- H2O added a feature to DriverlessAI so that it will write to STDOUT when running in a container -- that way the logs will get picked up by ICP's "built in" filebeat. Just pushed the latest H2O Chart to ==> https://github.com/ppc64le/charts

[alanchia67]

I have sync the repo from URL https://raw.githubusercontent.com/ppc64le/charts/master/repo/stable

It seems the repo didnt pickup the changes suggested by jstzsyf, as the mount volume failure persists.

[alanchia67]

Hi Bruce, any updates to my remarks on Jun 17? The repo I mentioned didnt pick up the changes and the filebeat sidecar is still required.