Remove deprecatedPublicIPs field

[thockin]

No description provided.

[k8s-reviewable]

This change is 

[xiangpengzhao]

@thockin Did you generate pkg/api/v1/types.generated.go correctly? It's an empty file in your PR and it leads to the tests failure.

[xiangpengzhao]

Please check this generated file:)

[xiangpengzhao]

Also change the name of $PUBLIC_IP to EXTERNAL_IP appears in this script?

[thockin]

@caesarxuchao I don't know what to do wrt staging updates. I ran update-staging-client (update-all) and it made a bunch of changes like the following, which seem wrong.

@@ -15,11 +15,24 @@ go_library(
     ],
     tags = ["automanaged"],
     deps = [
+        "//pkg/apis/certificates:go_default_library",
+        "//pkg/apis/certificates/v1beta1:go_default_library",
         "//vendor/k8s.io/apimachinery/pkg/api/errors:go_default_library",
         "//vendor/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
         "//vendor/k8s.io/apimachinery/pkg/labels:go_default_library",
-        "//vendor/k8s.io/client-go/pkg/apis/certificates:go_default_library",
-        "//vendor/k8s.io/client-go/pkg/apis/certificates/v1beta1:go_default_library",
         "//vendor/k8s.io/client-go/tools/cache:go_default_library",
     ],
 )

[caesarxuchao]

The Jenkins verification passed anyway because we ignore the BUILD file diffs: https://github.com/kubernetes/kubernetes/blob/master/staging/copy.sh#L220-L221. @thockin if you run hack/update-bazel.sh again, are those changes cancelled?

I don't know which script is managing the BUILD files in the staging folder, @mikedanese do you know?

[thockin]

I manually reverted the obviously broken BUILD changes. This is not a good developer experience. I made a simple API change, I had to run godep restore, manually revert BUILD files, and mess with my GOPATH. A casual contributor would have given up. *I* almost gave up.

…

On Mon, Apr 17, 2017 at 10:25 AM, Chao Xu ***@***.***> wrote: The Jenkins verification passed anyway because we ignore the BUILD file diffs: https://github.com/kubernetes/kubernetes/blob/master/ staging/copy.sh#L220-L221. @thockin <https://github.com/thockin> if you run hack/update-bazel.sh again, are those changes cancelled? I don't know which script is managing the BUILD files in the staging folder, @mikedanese <https://github.com/mikedanese> do you know? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#44519 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFVgVHZvY0vtXa4dLYA1KXKJ3LLgnFKAks5rw6B7gaJpZM4M-JcP> .

[bowei]

/lgtm

[k8s-github-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: bowei, thockin

Needs approval from an approver in each of these OWNERS Files:

• OWNERS [thockin]

You can indicate your approval by writing /approve in a comment
You can cancel your approval by writing /approve cancel in a comment

[mikedanese]

Gahh, that's a bug in how ./staging/copy.sh handles BUILD files, ./hack/update-bazel.sh is canonical. I'll take a look.

[caesarxuchao]

@mikedanese it seems the problem is exposed by #40777

[thockin]

Chao: the staging scripts assume they can call godep-restore, but without a clean GOPATH that can fail in a hundred different ways, leaving the user confused and frustrated. It really needs to make a temporary space, change GOPATH, restore into that, then process. It sucks. We should not be asking users to do this.

…

On Mon, Apr 17, 2017 at 10:38 AM, Mike Danese ***@***.***> wrote: Gahh, that's a bug in how ./staging/copy.sh handles BUILD files, ./hack/update-bazel.sh is canonical. I'll take a look. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#44519 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFVgVJCx5dOkL-xeELaDOgj4u6rH-V0Uks5rw6OXgaJpZM4M-JcP> .

[k8s-github-robot]

Automatic merge from submit-queue

[caesarxuchao]

@thockin the update-staging-client-go.sh verifies the working dir is clean before proceeding. Or do you mean other things by "clean GOPATH"?

It really needs to make a temporary space, change GOPATH, restore into that, then process

We had considered this, but it will take too long to run.

[thockin]

On Mon, Apr 17, 2017 at 12:44 PM, Chao Xu ***@***.***> wrote: @thockin the update-staging-client-go.sh verifies the working dir is clean before proceeding. Or do you mean other things by "clean GOPATH"?

If you have github.com/foobar in your own GOPATH, and then you do a `godep restore` which includes that same package at a different rev it can fail in a dozen different ways. What I had to do to make it pass was: mkdir _gopath GOPATH=`pwd`/_gopath ./hack/godep-restore.sh GOPATH=`pwd`/_gopath:$GOPATH ./hack/update-staging-client.sh

It really needs to make a temporary space, change GOPATH, restore into that, then process We had considered this, but it will take too long to run.

The alternative is that it fails in hard to comprehend ways. We should not be making users do this at all.

[caesarxuchao]

If you have github.com/foobar in your own GOPATH, and then you do a
godep restore which includes that same package at a different rev it
can fail in a dozen different ways.

godep restore fails in such an environment regardless of the staging folder. Currently hack/update-staging-client.sh and hack/update-staging-godeps.sh requires running godep restore. I target to eliminate the need for update-staging-client.sh this quarter. For update-staging-godeps.sh, I think @sttts once had a version of that script that operates in a temporary space, we can add it back as another alternative to running hack/godep-restore.sh, if it's easy.

In general, i think we should focus on eliminating the staging folder, instead of building more tools around it.