Files
homelab-frontend/internal/serviceadapter/memory_adapter.go
T
rockandAdmin Bot ef5e736673
CI / CI (push) Successful in 3m20s
feat(cag): CAG integration — memory adapter, warmup semaphore, cache-aware routing (#47)
## Summary

Three issues in one PR. Implements Cache-Augmented Generation (CAG) support in the api-gw.

---

### #44 — Memory ServiceAdapter: standing-context resource

- Code-registers `memory` adapter with `standing-context` resource (`POST /memory/standing-context`)
- Fixes dispatcher bug: method-level `Auth.Capability` override now respected (was always using verb-derived capability)
- Auth: `memory:read` required via method-level override; `memory:write` for ingest
- `MEMORY_UPSTREAM_URL` env configures upstream

### #45 — Continuous batching: warmup deprioritisation + per-backend semaphore

- In-memory bounded queue per backend: `VLLM_MAX_QUEUE_DEPTH` slots (default 20)
- Concurrency limiter: `VLLM_MAX_NUM_SEQS` slots (default 4)
- `X-Poimen-Warmup: true` → sleep `WARMUP_DELAY_MS` (default 500ms) before queue entry, strip header, add `X-Priority: low`
- Queue full → 503 immediately, no goroutine pile-up
- Metrics: `warmup_requests_delayed_total`, `queue_full_requests_total` at `/metrics`

### #46 — Cache-aware routing: rendezvous hash + prefix cache hit logging

- `addresses: [...]` YAML field for multi-backend models (backward-compat with `address`)
- `X-Poimen-Cache-Key` header → rendezvous-hash backend selection (same key = same GPU)
- No cache key → round-robin
- Failed backend excluded 30s; next-highest-score fallback
- `vllm-num-cached-tokens` response header captured → `vllm_prefix_cache_hit_tokens_total` counter + structured log

---

## Verification

```
go build ./...    ✅
go vet ./...      ✅
go test -race ./... ✅ (all packages, 0 races)
```

## Reviewer verdicts

- reviewer-sre: SIGN-OFF
- reviewer-quality: SIGN-OFF

## Non-blocking findings (🟡🔵 only)

- `getOrCreateTransport` not goroutine-safe by design; pre-population invariant documented
- `time.Sleep(warmupDelay)` holds goroutine; context-cancel alternative left as TODO
- OTel trace_id not propagated into CAG log events (existing gap)

Closes #44 Closes #45 Closes #46

---------

Co-authored-by: Admin Bot <[email protected]>
Reviewed-on: #47
2026-09-27 11:16:40 +00:00

68 lines
1.7 KiB
Go

package serviceadapter
import "os"
// NewMemoryAdapter creates a code-registered ServiceAdapter for the memory service.
// Takes precedence over any config-file loaded memory adapter (main.go skips config
// adapters when a code-registered one with the same name exists).
// Upstream URL configured via MEMORY_UPSTREAM_URL env var.
func NewMemoryAdapter() *ServiceAdapter {
upstreamURL := os.Getenv("MEMORY_UPSTREAM_URL")
if upstreamURL == "" {
upstreamURL = "http://poimen-memory.poimen.svc.cluster.local:8080"
}
return &ServiceAdapter{
Namespace: "poimen",
ServiceName: "memory",
Spec: *GetMemorySpec(upstreamURL),
}
}
// GetMemorySpec returns the ServiceAdapter spec for the memory service.
func GetMemorySpec(upstreamURL string) *Spec {
return &Spec{
ServiceName: "memory",
Upstream: Upstream{
URL: upstreamURL,
TimeoutSeconds: 30,
},
Auth: Auth{
Required: true,
Capability: "memory:read",
},
Resources: []Resource{
{
Name: "ingest",
Methods: []Method{{
Verb: "POST",
UpstreamPath: "/memory/ingest",
Auth: &Auth{Required: true, Capability: "memory:write"},
}},
},
{
Name: "query",
Methods: []Method{{
Verb: "POST",
UpstreamPath: "/memory/query",
Auth: &Auth{Required: true, Capability: "memory:read"},
}},
},
{
Name: "skills",
Methods: []Method{{
Verb: "GET",
UpstreamPath: "/memory/skills",
}},
},
{
Name: "standing-context",
Methods: []Method{{
Verb: "POST",
UpstreamPath: "/memory/standing-context",
Auth: &Auth{Required: true, Capability: "memory:read"},
}},
},
},
}
}