feat: minimal mlab-ns compatibility shim

[bassosimone]

We implement a minimal mlab-ns compatibility shim.

The service has been EOL for a long time now. We recently reached out to integrators and it seems there are still some that need a small compatibility shim to avoid breaking legacy applications.

Their migration is fully underway, but there needs to be time to migrate a fleet of legacy clients, which are using the output of mlab-ns for archival purposes.

Accordingly, this pull request implements just the minimum functionality to avoid breaking those applications. Yet, note that the returned response is not actionable anymore for running ndt5 tests (and it can't be used for running ndt7 tests either).

The complete plan is this:

1. merge this diff

2. deploy

3. use dispatch.yaml to route to locate

4. discontinue mlab-ns

5. wait for the legacy applications queue to drain

6. circle back and zap this diff

This is probably the last chapter in mlab-ns glorious life.

See #185.

Adapted from #184.

---

This change is 

[bassosimone]

I am now looking into the failures. Let's first deal with the linting failures:

  Error: handler/mlabnscompat.go:252:5: sloppyLen: len(v2Result.Results) <= 0 can be len(v2Result.Results) == 0 (gocritic)
  	if len(v2Result.Results) <= 0 || v2Result.Results[0].Location == nil {
  	   ^
  Error: handler/mlabnscompat_test.go:93:16: httpNoBody: http.NoBody should be preferred to the nil request body (gocritic)
  			outerReq := httptest.NewRequest(tt.method, tt.url, nil)
  			            ^
  Error: handler/mlabnscompat_test.go:364:16: httpNoBody: http.NoBody should be preferred to the nil request body (gocritic)
  			req, err := http.NewRequest(http.MethodGet, srv.URL+tt.path, nil)
  			            ^
  3 issues:
  * gocritic: 3

I agree with the second and the third. The first one, TBH, is a matter of code style and I am not super happy about this level of exaggerated linting for arguably more correct code (<= 0 excludes the whole domain of invalid values and len returns int, so, IMHO that's just a good coding practice across languages).

With respect to the tests proper, the failure is this:

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOROOT
/opt/hostedtoolcache/go/1.25.7/x64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOPATH
/home/runner/go
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOBIN

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOTMPDIR

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOTOOLDIR
/opt/hostedtoolcache/go/1.25.7/x64/pkg/tool/linux_amd64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOOS
linux
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOARCH
amd64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOCACHE
/home/runner/.cache/go-build
/home/runner/work/_actions/shogo82148/actions-goveralls/v1/bin/goveralls_linux_amd64 -service=github -coverprofile=coverage.out
bad response status from coveralls: 530
error code: 1016
Error: Error: The process '/home/runner/work/_actions/shogo82148/actions-goveralls/v1/bin/goveralls_linux_amd64' failed with exit code 1

and it happens while uploading to coveralls. So, this also needs to be investigated.

[bassosimone]

I am now looking into the failures. Let's first deal with the linting failures:

  Error: handler/mlabnscompat.go:252:5: sloppyLen: len(v2Result.Results) <= 0 can be len(v2Result.Results) == 0 (gocritic)
  	if len(v2Result.Results) <= 0 || v2Result.Results[0].Location == nil {
  	   ^
  Error: handler/mlabnscompat_test.go:93:16: httpNoBody: http.NoBody should be preferred to the nil request body (gocritic)
  			outerReq := httptest.NewRequest(tt.method, tt.url, nil)
  			            ^
  Error: handler/mlabnscompat_test.go:364:16: httpNoBody: http.NoBody should be preferred to the nil request body (gocritic)
  			req, err := http.NewRequest(http.MethodGet, srv.URL+tt.path, nil)
  			            ^
  3 issues:
  * gocritic: 3

I agree with the second and the third. The first one, TBH, is a matter of code style and I am not super happy about this level of exaggerated linting for arguably more correct code (<= 0 excludes the whole domain of invalid values and len returns int, so, IMHO that's just a good coding practice across languages).

Fixed the second and the third one. Disable sloppyLen. We're already using staticcheck, which catches the real useful cases len(x) >= 0 and len(x) < 0 already. The len(x) <= 0 is valid code and I actually not only like it more but have trouble not using this pattern (maybe C++ habit, IDK, but being defensive is good mental habit).

[bassosimone]

With respect to the tests proper, the failure is this:

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOROOT
/opt/hostedtoolcache/go/1.25.7/x64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOPATH
/home/runner/go
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOBIN

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOTMPDIR

/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOTOOLDIR
/opt/hostedtoolcache/go/1.25.7/x64/pkg/tool/linux_amd64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOOS
linux
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOARCH
amd64
/opt/hostedtoolcache/go/1.25.7/x64/bin/go env GOCACHE
/home/runner/.cache/go-build
/home/runner/work/_actions/shogo82148/actions-goveralls/v1/bin/goveralls_linux_amd64 -service=github -coverprofile=coverage.out
bad response status from coveralls: 530
error code: 1016
Error: Error: The process '/home/runner/work/_actions/shogo82148/actions-goveralls/v1/bin/goveralls_linux_amd64' failed with exit code 1

and it happens while uploading to coveralls. So, this also needs to be investigated.

This is a coveralls outage and there's not much that we can do but wait 🙄

[bassosimone]

@robertodauria coveralls is up and running again! Please, take a look! 🙏

[coveralls]

Pull Request Test Coverage Report for Build 22629741708

Details

• 114 of 123 (92.68%) changed or added relevant lines in 3 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.4%) to 82.629%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
locate.go	0	9	0.0%

Totals
Change from base Build 22571383483:	0.4%
Covered Lines:	2502
Relevant Lines:	3028

---

💛 - Coveralls

[robertodauria]

LGTM with a nit.