cmd/compress: do not compress Vultr images

[lucab]

This skips the compression step for Vultr images, as the cloud import
API expects a raw disk image and does not accept compressed content.

Ref: coreos/fedora-coreos-docs#71 (comment)

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: lucab

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [lucab]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[dustymabe]

how much does this expand the actual storage we'll be using? In the docs could we just add a -d decompress option to coreos-installer download ?

[lucab]

In the docs could we just add a -d decompress option to coreos-installer download ?

I had the same idea, until I discovered that the Vultr API allows uploading images from public URLs, but not pushing a local file (i.e. one cannot fetch+decompress+upload).

[dustymabe]

In the docs could we just add a -d decompress option to coreos-installer download ?

I had the same idea, until I discovered that the Vultr API allows uploading images from public URLs, but not pushing a local file (i.e. one cannot fetch+decompress+upload).

foo - that sounds like a request we should be able to take to them to be able to handle a compressed artifact.

[dustymabe]

foo - that sounds like a request we should be able to take to them to be able to handle a compressed artifact.

but for now, i guess this is best

[lucab]

Yeah, I'd also like to be able to specify a digest for their platform to check after fetching but before decompressing/importing. Let's hold this for a few days and see if they are open to that, it would help sparing some disk space and bandwidth.

/hold

[lucab]

As per coreos/fedora-coreos-tracker#355 (comment) the API is not going to gain those features right now. How do we want to proceed here?

[dustymabe]

As per coreos/fedora-coreos-tracker#355 (comment) the API is not going to gain those features right now. How do we want to proceed here?

I'd really like to not store the images uncompressed. Some questions/suggestions:

• Is there a public ticket/issue tracking such a feature request (i.e. support for decompression on the server side)?
• Since the user already has access to vultr and vultr has object storage, could we add steps to the docs to download/decompress, upload to object storage, create image using API, delete from object storage?

It's a bit backwards, but the steps would cover everything and wouldn't require the user to set up their own HTTPS server.

[dustymabe]

I'd really like to not store the images uncompressed.

But... I'm not unwilling to be convinced otherwise. I'm just trying to suggest some alternatives until the server side is in a better state.

[lucab]

It's a bit backwards, but the steps would cover everything and wouldn't require the user to set up their own HTTPS server.

I just tried that and the flow is a bit longer but not too complex. We can keep the image compressed here, I'll update the docs adding the "decompress and upload to bucket" steps.

[dustymabe]

Thanks Luca!