kpromo: set git clone depth for pr subcommand to prevent downloading the whole repository

[chrischdi]

What type of PR is this?

/kind feature

What this PR does / why we need it:

This PR adds git clone options for cloning the kubernetes/k8s.io repository to only clone a depth of 1.

The kubernetes/k8s.io did grow in the past to a big (1,4GB to download) repository:

❯ git clone https://github.com/kubernetes/k8s.io.git
Cloning into 'k8s.io'...
remote: Enumerating objects: 211546, done.
remote: Counting objects: 100% (1397/1397), done.
remote: Compressing objects: 100% (881/881), done.
remote: Total 211546 (delta 428), reused 1294 (delta 359), pack-reused 210149 (from 1)
Receiving objects: 100% (211546/211546), 1.27 GiB | 6.65 MiB/s, done.
Resolving deltas: 100% (122622/122622), done.
Updating files: 100% (15059/15059), done.
❯ du -h -d 0 k8s.io
1.4G    k8s.io

This change should reduce the downloaded size from 1.27 GiB to currently ~29.29 MiB instead.

❯ git clone --depth 1 https://github.com/kubernetes/k8s.io.git
Cloning into 'k8s.io'...
remote: Enumerating objects: 17520, done.
remote: Counting objects: 100% (17520/17520), done.
remote: Compressing objects: 100% (9760/9760), done.
remote: Total 17520 (delta 8344), reused 14982 (delta 7181), pack-reused 0 (from 0)
Receiving objects: 100% (17520/17520), 29.29 MiB | 4.14 MiB/s, done.
Resolving deltas: 100% (8344/8344), done.
Updating files: 100% (15059/15059), done.
❯ du -h -d 0 k8s.io
142M    k8s.io

So this reduces the amount of data which is downloaded when running kpromo pr ... and by that also reduce the time for maintainers to run kpromo.

Which issue(s) this PR fixes:

Special notes for your reviewer:

Does this PR introduce a user-facing change?

kpromo pr: Only clone the head commit to reduce amount of data to download

[xmudrii]

I agree, lately running kpromo pr takes way too much time.
/lgtm
/approve
/hold
@cpanato @saschagrunert do you want to take a look as well?

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: chrischdi, saschagrunert, xmudrii

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 [saschagrunert,xmudrii]

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

[saschagrunert]

The question is if we break something, but yeah let's give it a try.

[xmudrii]

That's why I tagged you and @cpanato. 😅 But given that this PR changes only the pr command, I think we can't break that much. And even if we do break it, we would probably realise it very quickly.

[xmudrii]

Let's give this a try!
/hold cancel

[BenTheElder]

We should consider --filter=blob:none instead, that's what we use in prow these days.
It doesn't break git describe (unlike depth=1 and similar) but it does significantly reduce clone size.

https://github.blog/open-source/git/get-up-to-speed-with-partial-clone-and-shallow-clone/

[chrischdi]

We should consider --filter=blob:none instead, that's what we use in prow these days. It doesn't break git describe (unlike depth=1 and similar) but it does significantly reduce clone size.

github.blog/open-source/git/get-up-to-speed-with-partial-clone-and-shallow-clone

Oh, this is neat!

I'll gonna try that and file a PR to change to that :-)

[chrischdi]

Unfortunately: --filter=blob:none is not yet supported in the used library: go-git/go-git#713

[BenTheElder]

Ah, we do clones via exec in prow ... it's kinda nice for debugging but does introduce an external runtime dependency ... (we use a container image with git)

... it does also mean you can use the canonical git implementation ...