Migrate to encoding/json/v2

[inteon]

Replaces the github.com/json-iterator/go dependency with encoding/json/v2.
Performance is not yet great (feel free to push improvements/ create new PRs based on this PR):

# based on f90a164c304e010637632ae24ab0f05d66851259
$ go test -benchmem -run=^$ -bench ^BenchmarkFieldSet/serialize.*$ sigs.k8s.io/structured-merge-diff/v6/fieldpath -count=6 > pr.txt
$ go test -benchmem -run=^$ -bench ^BenchmarkFieldSet/serialize.*$ sigs.k8s.io/structured-merge-diff/v6/fieldpath -count=6 > master.txt
$ benchstat master.txt pr.txt

goos: linux
goarch: amd64
pkg: sigs.k8s.io/structured-merge-diff/v6/fieldpath
cpu: Intel(R) Core(TM) Ultra 7 165H
                            │  master.txt  │                pr.txt                │
                            │    sec/op    │    sec/op     vs base                │
FieldSet/serialize-20-8       3.647µ ±  3%    7.915µ ± 1%  +117.04% (p=0.002 n=6)
FieldSet/deserialize-20-8     11.07µ ±  7%    15.64µ ± 2%   +41.22% (p=0.002 n=6)
FieldSet/serialize-50-8       10.82µ ±  1%    21.85µ ± 2%  +102.00% (p=0.002 n=6)
FieldSet/deserialize-50-8     29.76µ ± 36%    43.60µ ± 2%   +46.48% (p=0.002 n=6)
FieldSet/serialize-100-8      38.58µ ±  4%    78.64µ ± 5%  +103.87% (p=0.002 n=6)
FieldSet/deserialize-100-8    83.91µ ±  2%   143.50µ ± 1%   +71.01% (p=0.002 n=6)
FieldSet/serialize-500-8      228.9µ ± 44%    420.8µ ± 6%   +83.84% (p=0.002 n=6)
FieldSet/deserialize-500-8    423.9µ ±  4%    699.2µ ± 1%   +64.95% (p=0.002 n=6)
FieldSet/serialize-1000-8     527.7µ ±  6%    929.0µ ± 2%   +76.04% (p=0.002 n=6)
FieldSet/deserialize-1000-8   979.2µ ± 33%   1578.9µ ± 5%   +61.24% (p=0.002 n=6)
geomean                       67.99µ          119.1µ        +75.18%

                            │  master.txt  │                pr.txt                │
                            │     B/op     │     B/op       vs base               │
FieldSet/serialize-20-8       1.543Ki ± 0%    2.754Ki ± 0%  +78.48% (p=0.002 n=6)
FieldSet/deserialize-20-8     9.761Ki ± 0%    8.269Ki ± 0%  -15.28% (p=0.002 n=6)
FieldSet/serialize-50-8       4.466Ki ± 0%    7.451Ki ± 0%  +66.83% (p=0.002 n=6)
FieldSet/deserialize-50-8     19.87Ki ± 0%    23.29Ki ± 0%  +17.22% (p=0.002 n=6)
FieldSet/serialize-100-8      15.73Ki ± 0%    25.00Ki ± 0%  +58.89% (p=0.002 n=6)
FieldSet/deserialize-100-8    59.21Ki ± 0%    80.49Ki ± 0%  +35.95% (p=0.002 n=6)
FieldSet/serialize-500-8      73.77Ki ± 0%   123.98Ki ± 0%  +68.06% (p=0.002 n=6)
FieldSet/deserialize-500-8    269.9Ki ± 0%    394.6Ki ± 0%  +46.20% (p=0.002 n=6)
FieldSet/serialize-1000-8     154.4Ki ± 1%    269.9Ki ± 0%  +74.84% (p=0.002 n=6)
FieldSet/deserialize-1000-8   593.5Ki ± 0%    870.3Ki ± 0%  +46.63% (p=0.002 n=6)
geomean                       34.33Ki         49.67Ki       +44.70%

                            │ master.txt  │               pr.txt                │
                            │  allocs/op  │  allocs/op   vs base                │
FieldSet/serialize-20-8        8.000 ± 0%   28.000 ± 0%  +250.00% (p=0.002 n=6)
FieldSet/deserialize-20-8      232.0 ± 0%    254.0 ± 0%    +9.48% (p=0.002 n=6)
FieldSet/serialize-50-8        14.00 ± 0%    74.00 ± 0%  +428.57% (p=0.002 n=6)
FieldSet/deserialize-50-8      655.0 ± 0%    739.0 ± 0%   +12.82% (p=0.002 n=6)
FieldSet/serialize-100-8       39.00 ± 0%   249.00 ± 0%  +538.46% (p=0.002 n=6)
FieldSet/deserialize-100-8    2.305k ± 0%   2.595k ± 0%   +12.60% (p=0.002 n=6)
FieldSet/serialize-500-8       185.0 ± 0%   1234.0 ± 0%  +567.03% (p=0.002 n=6)
FieldSet/deserialize-500-8    11.46k ± 0%   13.00k ± 0%   +13.45% (p=0.002 n=6)
FieldSet/serialize-1000-8      393.0 ± 0%   2662.0 ± 0%  +577.35% (p=0.002 n=6)
FieldSet/deserialize-1000-8   25.01k ± 0%   28.39k ± 0%   +13.52% (p=0.002 n=6)
geomean                        355.1         888.0       +150.06%

closes #202

[dims]

xref: kubernetes/kubernetes#132312

[dims]

For the pull-structured-merge-diff-test failure, please add this to fix the error @inteon

diff --git a/internal/cli/main_test.go b/internal/cli/main_test.go
index 3e409ede0673..8beed5c6cf55 100644
--- a/internal/cli/main_test.go
+++ b/internal/cli/main_test.go
@@ -21,6 +21,7 @@ import (
        "encoding/json"
        "io/ioutil"
        "path/filepath"
+       "strings"
        "testing"
 )

@@ -135,7 +136,7 @@ func (tt *testCase) checkOutput(t *testing.T, got []byte) {
                t.Fatalf("couldn't read expected output %q: %v", tt.expectedOutputPath, err)
        }

-       if a, e := string(got), string(want); a != e {
+       if a, e := strings.TrimSpace(string(got)), strings.TrimSpace(string(want)); a != e {
                t.Errorf("output didn't match expected output: got:\n%v\nwanted:\n%v\n", a, e)
        }
 }

[inteon]

For the pull-structured-merge-diff-test failure, please add this to fix the error @inteon
...

I fixed the test failure.

[dims]

/assign @BenTheElder @liggitt

[liggitt]

/assign @jpbetz
who is the primary apimachinery approver on this bit and was deeply involved in the initial performance-driven use of json-iterator in these bits

[liggitt]

For the pull-structured-merge-diff-test failure, please add this to fix the error @inteon

I suspect using a json marshal function (like MarshalWrite) that doesn't append a newline would be a more efficient way to accomplish that

[liggitt]

is rawKey.String() the same as decoding to a string, in terms of interpreting escape sequences, etc?

[liggitt]

curious if it's ok to drop these... were they added to try to catch a specific issue?

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: inteon
Once this PR has been reviewed and has the lgtm label, please ask for approval from jpbetz. For more information see the Code Review Process.

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

Needs approval from an approver in each of these files:

• OWNERS

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

[liggitt]

The .../deserialize... benchmarks actually don't look terrible now... I'd be willing to accept that performance drop in pursuit of correctness / safety.

The serialize benchmarks still look pretty rough. Need to see what we can improve there.

[liggitt]

did you run the full set of benchmarks to see how we looked across all of them?

[liggitt]

Thanks for the updates, how are the overall benchmarks looking (not just the subset in the description)?

As you were adjusting the implementation, were there any unit tests it would make sense to add to catch edges the previous implementations handled we want to ensure the new one does as well? I'm thinking specifically of things like:

• handling of extra data in the input bytes/buffer when decoding/deserializing (e.g. "somevalue"extrastuff or {"key":"value"}extrastuff)
• handling of special characters in strings that need escaping where the raw bytes would not be the same as the decoded or encoded/escaped bytes
• handling of ignorable whitespace when decoding

[jpbetz]

First off- it's amazing to see this happening and the benchmarks are VERY promising. Thanks @inteon!

To get this to the finish line, and merge, what should our criteria be?

I chatted offline with @liggitt briefly and some of the criteria we discussed was:

• Golang releases a stable json/v2 (The alternative would be to add an internal copy of json-experiment to this repo like kube-openapi has, but I don't know if it's worth it given how close json/v2 is to stable)
• github.com/kubernetes/kubernetes CI test stability is not negatively impacted
• This passes a scale test (SIG instrumentation)
• We are confident on the correctness (triple check the implementation, shore up with additional functional tests)

Intuitively, it seems like the deserialization is already sufficiently fast. I suspect we need to optimize serialization a bit further since we serialize managed fields on all updates (not just patches). That said, I'm willing to be data driven here. If we can show downstream scale and performance is acceptable, I'm willing to accept a higher serialization perf regression in order to migrate to json/v2.

Thoughts, concerns?