Files
coder/coderd/aibridged.go
T
Danny Kopping ce94d42e19 feat: fetch providers over DRPC (#26650)
Closes [AIGOV-455](https://linear.app/codercom/issue/AIGOV-455/extend-drpc-with-buildproviders).

## Why

The AI Gateway (`aibridged`) is being split into a standalone process that must not touch the database. `coderd` stays the source of truth and seeds the `ai_providers` / `ai_provider_keys` tables from the environment. This PR adds a DRPC call so the gateway fetches provider config from `coderd` instead of reading the DB, for both the embedded and standalone daemons.

## What

- **Proto:** new `ProviderConfigurator` service with a unary `GetAIProviders` RPC, plus `AIProvider` / `AIProviderBedrock` messages. `CurrentMinor` bumped to 1 (additive).
- **Server (`coderd/aibridgedserver`):** `GetAIProviders` runs a read-only `InTx` under `LockIDAIProvidersEnvSeed` so it never returns a mid-seed snapshot, reads providers (incl. disabled) plus keys for enabled ones, and maps to proto under `dbauthz.AsAIBridged`. Unmappable rows are skipped and logged; plaintext keys and Bedrock secrets are never logged.
- **Client:** `DRPCProviderConfiguratorClient` wired into the client union, `dialer.go`, and `CreateInMemoryAIBridgeServer`.
- **cli:** `BuildProvidersFromProto` maps the response through the existing DB-neutral `buildProvider`. A shared `poolRPCReloader` does the fetch/build/replace for both daemons: the embedded daemon reloads on every `ai_providers` change and fails startup if it cannot subscribe; the standalone gateway drives the same reloader once at startup, retrying until success and staying interruptible.
- **Dead code removed:** `BuildProvidersFromConfig`, `ProvidersFromConfig`, `AIProviderFromConfig`, and the DB-read `BuildProviders` path.
2026-06-29 13:34:58 +02:00

113 lines
4.3 KiB
Go

package coderd
import (
"context"
"errors"
"io"
"net/http"
"storj.io/drpc/drpcmux"
"storj.io/drpc/drpcserver"
"cdr.dev/slog/v3"
agplaibridge "github.com/coder/coder/v2/coderd/aibridge"
"github.com/coder/coder/v2/coderd/aibridged"
aibridgedproto "github.com/coder/coder/v2/coderd/aibridged/proto"
"github.com/coder/coder/v2/coderd/aibridgedserver"
"github.com/coder/coder/v2/coderd/tracing"
"github.com/coder/coder/v2/codersdk/drpcsdk"
)
// AIGatewayHandler returns the in-memory AI Gateway HTTP handler
// set by [API.RegisterInMemoryAIBridgedHTTPHandler], or nil if the daemon
// has not been wired in. Callers must apply their own [http.StripPrefix]
// for the route prefix they are mounting under.
func (api *API) AIGatewayHandler() http.Handler {
return api.aiGatewayHandler
}
// RegisterInMemoryAIBridgedHTTPHandler mounts [aibridged.Server]'s HTTP router onto
// [API]'s router, so that requests to aibridged will be relayed from Coder's API server
// to the in-memory aibridged.
//
// This also registers an in-process [agplaibridge.TransportFactory] so that
// chatd can route coder-agent LLM traffic through aibridge without crossing
// the HTTP route. No license entitlement gate is applied at the factory layer:
// the entitlement check stays on the HTTP route for external callers, while
// in-process coder-agent traffic is the explicit carve-out.
func (api *API) RegisterInMemoryAIBridgedHTTPHandler(srv http.Handler) {
if srv == nil {
panic("aibridged cannot be nil")
}
api.aiGatewayHandler = srv
factory := aibridged.NewTransportFactory(http.StripPrefix(agplaibridge.AIGatewayRootPath, srv))
var asInterface agplaibridge.TransportFactory = factory
api.AIBridgeTransportFactory.Store(&asInterface)
}
// CreateInMemoryAIBridgeServer creates a [aibridged.DRPCServer] and returns a
// [aibridged.DRPCClient] to it, connected over an in-memory transport.
// This server is responsible for all the Coder-specific functionality that aibridged
// requires such as persistence and retrieving configuration.
func (api *API) CreateInMemoryAIBridgeServer(dialCtx context.Context) (client aibridged.DRPCClient, err error) {
// TODO(dannyk): implement options.
// TODO(dannyk): implement tracing.
// TODO(dannyk): implement API versioning.
clientSession, serverSession := drpcsdk.MemTransportPipe()
defer func() {
if err != nil {
_ = clientSession.Close()
_ = serverSession.Close()
}
}()
mux := drpcmux.New()
srv, err := aibridgedserver.NewServer(api.ctx, api.Database, api.Logger.Named("aibridgedserver"),
api.AccessURL.String(), api.DeploymentValues.AI.BridgeConfig, api.ExternalAuthConfigs, api.Experiments, api.AISeatTracker)
if err != nil {
return nil, err
}
if err := aibridgedserver.Register(mux, srv); err != nil {
return nil, err
}
server := drpcserver.NewWithOptions(&tracing.DRPCHandler{Handler: mux},
drpcserver.Options{
Manager: drpcsdk.DefaultDRPCOptions(nil),
Log: func(err error) {
if errors.Is(err, io.EOF) {
return
}
api.Logger.Debug(dialCtx, "aibridged drpc server error", slog.Error(err))
},
},
)
// in-mem pipes aren't technically "websockets" but they have the same properties as far as the
// API is concerned: they are long-lived connections that we need to close before completing
// shutdown of the API.
api.WebsocketWaitMutex.Lock()
api.WebsocketWaitGroup.Add(1)
api.WebsocketWaitMutex.Unlock()
go func() {
defer api.WebsocketWaitGroup.Done()
// Here we pass the background context, since we want the server to keep serving until the
// client hangs up. The aibridged is local, in-mem, so there isn't a danger of losing contact with it and
// having a dead connection we don't know the status of.
err := server.Serve(context.Background(), serverSession)
api.Logger.Info(dialCtx, "aibridge daemon disconnected", slog.Error(err))
// Close the sessions, so we don't leak goroutines serving them.
_ = clientSession.Close()
_ = serverSession.Close()
}()
return &aibridged.Client{
Conn: clientSession,
DRPCRecorderClient: aibridgedproto.NewDRPCRecorderClient(clientSession),
DRPCMCPConfiguratorClient: aibridgedproto.NewDRPCMCPConfiguratorClient(clientSession),
DRPCAuthorizerClient: aibridgedproto.NewDRPCAuthorizerClient(clientSession),
DRPCProviderConfiguratorClient: aibridgedproto.NewDRPCProviderConfiguratorClient(clientSession),
}, nil
}