happyDomain/internal/checker/observation_test.go
Pierre-Olivier Mercier 33702d3ee4 checkers: introduce checker subsystem foundation
Add the checker-sdk-go dependency and build the core checker
infrastructure:
- Domain model types: CheckTarget, CheckPlan, Execution,
  CheckEvaluation, CheckerDefinition, CheckerOptions,
  ObservationSnapshot, and associated interfaces
- Observation collection engine with concurrent per-key gathering
- Checker and observation provider registries (wrapping checker-sdk-go)
- WorstStatusAggregator for combining rule evaluation results
- AutoFill constants for context-driven option resolution
- Config option and CLI flag for max observation concurrency
- Revised error types for the checker domain
2026-04-10 16:53:52 +07:00

180 lines
5.2 KiB
Go

// This file is part of the happyDomain (R) project.
// Copyright (c) 2020-2026 happyDomain
// Authors: Pierre-Olivier Mercier, et al.
//
// This program is offered under a commercial and under the AGPL license.
// For commercial licensing, contact us at <contact@happydomain.org>.
//
// For AGPL licensing:
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
package checker
import (
"context"
"sync"
"sync/atomic"
"testing"
"time"
"git.happydns.org/happyDomain/model"
)
// blockingProvider is an ObservationProvider whose Collect blocks on the
// release channel until the test signals it. It records how many concurrent
// Collect calls are in flight at any moment.
type blockingProvider struct {
key happydns.ObservationKey
release chan struct{}
calls int32
maxCalls int32
}
func (b *blockingProvider) Key() happydns.ObservationKey { return b.key }
func (b *blockingProvider) Collect(ctx context.Context, _ happydns.CheckerOptions) (any, error) {
atomic.AddInt32(&b.calls, 1)
defer atomic.AddInt32(&b.calls, -1)
for {
current := atomic.LoadInt32(&b.calls)
max := atomic.LoadInt32(&b.maxCalls)
if current > max {
if atomic.CompareAndSwapInt32(&b.maxCalls, max, current) {
break
}
continue
}
break
}
select {
case <-b.release:
return map[string]string{string(b.key): "ok"}, nil
case <-ctx.Done():
return nil, ctx.Err()
}
}
// TestObservationContext_ConcurrentDifferentKeys verifies that two Get calls
// for distinct observation keys can run their Collect concurrently — i.e.
// the per-context lock is not held across provider.Collect.
func TestObservationContext_ConcurrentDifferentKeys(t *testing.T) {
release := make(chan struct{})
defer close(release)
pa := &blockingProvider{key: happydns.ObservationKey("test-a"), release: release}
pb := &blockingProvider{key: happydns.ObservationKey("test-b"), release: release}
oc := NewObservationContext(happydns.CheckTarget{}, happydns.CheckerOptions{}, nil, 0)
oc.SetProviderOverride(pa.key, pa)
oc.SetProviderOverride(pb.key, pb)
ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
defer cancel()
var wg sync.WaitGroup
results := make([]error, 2)
for i, key := range []happydns.ObservationKey{pa.key, pb.key} {
wg.Add(1)
go func(idx int, k happydns.ObservationKey) {
defer wg.Done()
var dst map[string]string
results[idx] = oc.Get(ctx, k, &dst)
}(i, key)
}
// Wait until both providers are blocked inside Collect simultaneously.
deadline := time.Now().Add(2 * time.Second)
for time.Now().Before(deadline) {
if atomic.LoadInt32(&pa.calls) == 1 && atomic.LoadInt32(&pb.calls) == 1 {
break
}
time.Sleep(5 * time.Millisecond)
}
if a, b := atomic.LoadInt32(&pa.calls), atomic.LoadInt32(&pb.calls); a != 1 || b != 1 {
t.Fatalf("expected both providers to be collecting in parallel, got a=%d b=%d", a, b)
}
// Release both Collects and wait for the Get calls to return.
release <- struct{}{}
release <- struct{}{}
wg.Wait()
for i, err := range results {
if err != nil {
t.Errorf("Get %d returned error: %v", i, err)
}
}
}
// TestObservationContext_DedupesSameKey verifies that concurrent Get calls
// for the *same* key only invoke provider.Collect once.
func TestObservationContext_DedupesSameKey(t *testing.T) {
release := make(chan struct{})
var collectCount int32
prov := &countingProvider{
key: happydns.ObservationKey("test-dedup"),
release: release,
count: &collectCount,
}
oc := NewObservationContext(happydns.CheckTarget{}, happydns.CheckerOptions{}, nil, 0)
oc.SetProviderOverride(prov.key, prov)
ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
defer cancel()
const N = 8
var wg sync.WaitGroup
wg.Add(N)
for i := 0; i < N; i++ {
go func() {
defer wg.Done()
var dst map[string]string
if err := oc.Get(ctx, prov.key, &dst); err != nil {
t.Errorf("Get error: %v", err)
}
}()
}
// Wait for at least one collect to be in flight, then release it.
deadline := time.Now().Add(2 * time.Second)
for time.Now().Before(deadline) && atomic.LoadInt32(&collectCount) == 0 {
time.Sleep(5 * time.Millisecond)
}
close(release)
wg.Wait()
if got := atomic.LoadInt32(&collectCount); got != 1 {
t.Errorf("expected exactly 1 Collect call, got %d", got)
}
}
type countingProvider struct {
key happydns.ObservationKey
release chan struct{}
count *int32
}
func (c *countingProvider) Key() happydns.ObservationKey { return c.key }
func (c *countingProvider) Collect(ctx context.Context, _ happydns.CheckerOptions) (any, error) {
atomic.AddInt32(c.count, 1)
select {
case <-c.release:
return map[string]string{"k": "v"}, nil
case <-ctx.Done():
return nil, ctx.Err()
}
}