I have a Google Cloud Storage bucket and am experimenting with creating a website in Google Cloud Platform (using Python+Flask FWIW). For my initial experiment, I'm using gcsfuse to mount my GCS bucket as a filesystem:
https://github.com/GoogleCloudPlatform/gcsfuse
Following the directions I got it installed in my cloud shell and mounted it like so:
mkdir -p ~/mnt
gcsfuse <my-bucket> ~/mnt
This works, but gives me a strange (as in wildly inaccurate) view of my bucket. It shows some YYYY-MM directory structure (all empty) and a bunch of files at the top level (though far from all of them), but it doesn't look like what's really in my bucket. I can verify my actual data is still available by visiting my existing GCS-homed static website and wandering through the archived YYYY-MM directories by date and thread, viewing the messages they contain:
http://www.smontanaro.net/
(I know, it's ugly. That's one reason to use something more modern to produce a decent website.)
I am new to gcsfuse and Google Cloud Platform, so it's quite likely I'm doing something wrong. Any suggestions about where to look for clues would be much appreciated.
Edit: I upload files from my laptop using gsutil:
gsutil -m cp -I gs://<my-bucket>/path...
(It's not really big data stuff, less than 10gb. GCS was just an easy place to toss it and make it available through the web.)
Edit: I created a new bucket and tossed in a few files. I can gcsfuse mount it from a Cloud shell (for either my GCP project or GCS project) and see the files. I've been so far unsuccessful trying to mount it from my Linux laptop, getting:
SetUpBucket: OpenBucket: Bad credentials for bucket "<my-bucket>": permission denied
This when specifying a JSON key file I created for my GCP project.