4_Test_Container_for_Pipelines.md => constantly restarting docker image
nick-youngblut opened this issue · 1 comments
I created Cloud Compute VM with registry.hub.docker.com/lynnlangit/blastn used for Container image in the Deploy image VM config options. All other VM config options were set to default.
When I ssh onto the VM and run docker ps, I see that the container is constantly restarting every ~1 minute. The same occurs if I use a number of different Docker images (public or from my GCP Artifact Registry). The VM logs do not provide any insight into why the docker image is restarting. I've tried explicitly setting the container command to /bin/bash or used custom containers with either and ENTRYPOINT or CMD set to /bin/bash. No matter why I try, the container constantly restarts in the VM.
If I change the restart policy to never, then docker ps shows nothing. So, it appears that the container is either just running /bin/bash and then ending (and then restarting, with the always container restart policy) or the container is failing without even running /bin/bash. Again, the VM logs do not provide useful info on what is going on.
It would be great to have some info in the docs about how to use a VM in a container (via specifying the container in the VM setup) and docker attach.
In regards to using an OS-Optimized container and then pulling a Docker image after starting up the VM (as in the video), the OS-Optimized containers do not have gcloud or gsutil installed. apt-get is not available in the image, so gcloud cannot be installed that way. Installing gcloud via ./google-cloud-sdk/install.sh results in module not found errors.
Without availability of gsutil (or gcsfuse), the VM is rather limited, if one wants to use data from GCP Cloud Storage.