Bumping up etcd dep to fix data race

[hongchaodeng]

ref: #23694

What this PR does

• Bumping up the godep of etcd to fix data race in etcd watcher.

[k8s-github-robot]

Labelling this PR as size/XS

[k8s-bot]

GCE e2e build/test failed for commit 34908ac9b4145ab2992ef20d3794ca251c9396a4.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Please reference the list of currently known flakes when examining this failure. If you request a re-test, you must reference the issue describing the flake.

[k8s-bot]

GCE e2e build/test failed for commit eff39aa3cd8c60dd86850e965b91e431618cd7ef.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Please reference the list of currently known flakes when examining this failure. If you request a re-test, you must reference the issue describing the flake.

[k8s-bot]

GCE e2e build/test failed for commit 812e120a100c7f49cb7e479c43a9322825acbd38.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Please reference the list of currently known flakes when examining this failure. If you request a re-test, you must reference the issue describing the flake.

[hongchaodeng]

It seemed like etcd master dropped support for go 1.4

[eparis]

I assigned the PR to you. You should 'unassign' the PR when you are ready. Or ping me/re-assign to me.

[hongchaodeng]

Blocked on #22149

[hongchaodeng]

Rebased after the merged Go 1.6 PR.
I'm just bumping up the godep of etcd. But I'm adding the watcher commit here to test if the data race is fixed. Will remove it after the build pass.

/cc @wojtek-t @gmarek @xiang90

[k8s-bot]

GCE e2e build/test passed for commit 1c11fc65e021c275e05c7aa7978c3e8e499ca1a9.

• Build Log
• Test Artifacts
• Internal Jenkins Results

[k8s-github-robot]

The author of this PR is not in the whitelist for merge, can one of the admins add the 'ok-to-merge' label?

[k8s-bot]

GCE e2e build/test passed for commit f9af323.

• Build Log
• Test Artifacts
• Internal Jenkins Results

[gmarek]

@hongchaodeng please squash commits.

[Random-Liu]

/cc @ixdy

[ixdy]

Should probably be explicitly noted somewhere that this PR ratchets us into Go 1.6+, assuming I'm understanding correctly.

[Random-Liu]

@ixdy
This PR build failed because

Godeps/_workspace/src/github.com/coreos/etcd/client/cancelreq.go:13:
req.Cancel undefined (type *http.Request has no field or method Cancel)

The Cancel field is added after go 1.5 (See here), while it's not there in go 1.4 (See here).

[ixdy]

OK, though due to #20656 Kubernetes basically doesn't support go 1.5.

[ixdy]

My basic worry is thus: we just merged the Go 1.6 PR yesterday, and we've been seeing some failures in CI that we can't fully explain away yet. I'm hesitant to lock us into Go 1.5+ until we're sure that things are not broken and we're not going to roll back.

cc @kubernetes/goog-testing

[hongchaodeng]

Closing this to transfer discussion to #23915.

Once go version is resolved, will open up another PR.

[wojtek-t]

@ixdy - one point from me about Go 1.6 to note - we don't observe any performance issues with it, so from performance perspective the upgrade is fine.

[ixdy]

all Google-run Jenkins VMs should be using Go 1.6 now.