mirror of
https://github.com/hanzoai/ai.git
synced 2026-11-20 14:13:23 +00:00
hanzokv/go v9.22.0 renames its Go package from `redis` to `kv` (import path unchanged). Both cache-bus files imported it under an explicit `redis` alias; they now use `kv`. v9.22.0 ALSO dropped the redis:// URL scheme: ParseURL now accepts only kv://, kvs:// and unix://. The integration test pointed KV_URL at redis://<miniredis> and started failing for real (bus refused to parse the URL and fell back to per-pod TTL), so it now uses kv://. OPERATOR IMPACT: any deployment setting KV_URL=redis://... silently loses the fleet-wide cache-invalidation bus and degrades to per-pod TTL, because InitCacheBus fails soft by design. Move KV_URL to kv:// when this ships. Verified: go build ./object/... and go vet ./object/... clean; TestCacheBus_CrossPodInvalidation PASSES (full pub/sub loop).
59 lines
2.0 KiB
Go
59 lines
2.0 KiB
Go
package object
|
|
|
|
import (
|
|
"context"
|
|
"testing"
|
|
"time"
|
|
|
|
"github.com/alicebob/miniredis/v2"
|
|
kv "github.com/hanzokv/go/v9"
|
|
)
|
|
|
|
// TestCacheBus_CrossPodInvalidation proves the FULL pub/sub loop end-to-end against
|
|
// a real Redis-protocol broker (miniredis, in-process): a DIFFERENT pod's write,
|
|
// published to the invalidation channel, reaches THIS pod's subscriber and flushes
|
|
// its local org-settings cache — the sub-second cross-pod convergence KV_URL buys.
|
|
func TestCacheBus_CrossPodInvalidation(t *testing.T) {
|
|
mr := miniredis.RunT(t) // in-process Redis w/ pub-sub; auto-closed by t.Cleanup
|
|
|
|
// Bring up THIS pod's bus (subscriber) pointed at the broker. kv-go v9.22.0
|
|
// accepts ONLY the kv:// (and kvs://, unix://) scheme — redis:// was dropped.
|
|
t.Setenv("KV_URL", "kv://"+mr.Addr())
|
|
InitCacheBus()
|
|
t.Cleanup(func() {
|
|
cacheBusMu.Lock()
|
|
if cacheBusClient != nil {
|
|
_ = cacheBusClient.Close()
|
|
cacheBusClient = nil
|
|
}
|
|
cacheBusMu.Unlock()
|
|
})
|
|
|
|
// Seed this pod's cache as if it had served a read.
|
|
invalidateOrgSettingsCacheLocal()
|
|
orgSettingsCacheMu.Lock()
|
|
orgSettingsCache["acme"] = &orgSettingsCacheEntry{settings: &OrgSettings{Owner: "acme"}, fetchedAt: time.Now()}
|
|
orgSettingsCacheMu.Unlock()
|
|
|
|
// A SEPARATE pod publishes an org_settings write. Retry until it lands — pub/sub
|
|
// is lossy until this pod's subscriber has attached — or time out. A live pod
|
|
// subscribes at boot, long before any write, so it needs no retry in production.
|
|
pub := kv.NewClient(&kv.Options{Addr: mr.Addr()})
|
|
defer pub.Close()
|
|
|
|
deadline := time.Now().Add(5 * time.Second)
|
|
for {
|
|
_ = pub.Publish(context.Background(), cacheBusChannel, "org_settings").Err()
|
|
orgSettingsCacheMu.RLock()
|
|
n := len(orgSettingsCache)
|
|
orgSettingsCacheMu.RUnlock()
|
|
if n == 0 {
|
|
return // subscriber received the broadcast and flushed — loop proven
|
|
}
|
|
if time.Now().After(deadline) {
|
|
t.Fatal("cross-pod broadcast did not flush the local cache within 5s")
|
|
}
|
|
time.Sleep(20 * time.Millisecond)
|
|
}
|
|
}
|