Add PoST service to node (#5091)

## Motivation
Part of #5042

This PR adds the GRPC API for the PoST service and adds tests verifying it works correctly.

## Changes
- PoST service is downloaded with tests and builds, as well as included in the `Dockerfile`
- `PostSetupManager` now uses `zap` directly instead of the `go-spacemesh/log` wrapper
- `grpcserver.PostService` uses the PoST service API is added to gRPCs `PrivateListener` endpoint
  - temporary, will be changed to public + authentication in the future
- Any amount of PoST services can connect to the API but only the first calling `Register` will be able to successfully create a stream
  - in the future this will be extended to allow any number of PoST services to connect
  - a successful `Register` will create a `PostClient` and notify all services that are `postConnectionListener` about the service connecting (and disconnecting)
- TODO: bundled PoST service is started when the node starts 

## Test Plan
- new tests have been added for gRPC API

## TODO
<!-- This section should be removed when all items are complete -->
- [x] Explain motivation or link existing issue(s)
- [x] Test changes and document test plan
- [x] Update documentation as needed
- [x] Update [changelog](../CHANGELOG.md) as needed

CHANGELOG.md

@@ -4,6 +4,12 @@ See [RELEASE](./RELEASE.md) for workflow instructions.
 
 ## UNRELEASED
 
+### Upgrade information
+
+### Highlights
+
+### Features
+
 ### Improvements
 
 * [#5118](https://github.com/spacemeshos/go-spacemesh/pull/5118) reduce number of tortoise results returned after recovery.
@@ -12,6 +18,8 @@ this is hotfix for a bug introduced in v1.2.0. in rare conditions node may loop
 
 > 2023-10-02T15:28:14.002+0200 WARN fd68b.sync mesh failed to process layer from sync {"node_id": "fd68b9397572556c2f329f3e5af2faf23aef85dbbbb7e38447fae2f4ef38899f", "module": "sync", "sessionId": "29422935-68d6-47d1-87a8-02293aa181f3", "layer_id": 23104, "errmsg": "requested layer 8063 is before evicted 13102", "name": "sync"}
 
+* [#5091](https://github.com/spacemeshos/go-spacemesh/pull/5091) First stage of separating PoST from the node into its own service.
+
 ## v1.2.0
 
 ### Upgrade information
@@ -42,6 +50,7 @@ Support for old certificate sync protocol is dropped. This update is incompatibl
 * [#5067](https://github.com/spacemeshos/go-spacemesh/pull/5067) dbstat virtual table can be read periodically to collect table/index sizes.
 
 In order to enable provide following configuration:
+
 ```json
 "main": {
     "db-size-metering-interval": "10m"

Dockerfile

@@ -56,6 +56,7 @@ FROM linux AS spacemesh
 
 # Finally we copy the statically compiled Go binary.
 COPY --from=builder /src/build/go-spacemesh /bin/
+COPY --from=builder /src/build/service /bin/
 COPY --from=builder /src/build/libpost.so /bin/
 COPY --from=builder /src/build/gen-p2p-identity /bin/
 

Makefile

@@ -52,10 +52,10 @@ install:
 	go install honnef.co/go/tools/cmd/[email protected]
 .PHONY: install
 
-build: go-spacemesh get-profiler
+build: go-spacemesh get-profiler get-postrs-service
 .PHONY: build
 
-get-libs: get-postrs-lib
+get-libs: get-postrs-lib get-postrs-service
 .PHONY: get-libs
 
 get-profiler: get-postrs-profiler

Makefile-libs.Inc

@@ -50,22 +50,28 @@ else
 	endif
 endif
 
-POSTRS_SETUP_REV = 0.4.4
+POSTRS_SETUP_REV = 0.5.0-alpha2
 POSTRS_SETUP_ZIP = libpost-$(platform)-v$(POSTRS_SETUP_REV).zip
 POSTRS_SETUP_URL_ZIP ?= https://github.com/spacemeshos/post-rs/releases/download/v$(POSTRS_SETUP_REV)/$(POSTRS_SETUP_ZIP)
 POSTRS_PROFILER_ZIP = profiler-$(platform)-v$(POSTRS_SETUP_REV).zip
 POSTRS_PROFILER_URL ?= https://github.com/spacemeshos/post-rs/releases/download/v$(POSTRS_SETUP_REV)/$(POSTRS_PROFILER_ZIP)
+POSTRS_SERVICE_ZIP = service-$(platform)-v$(POSTRS_SETUP_REV).zip
+POSTRS_SERVICE_URL ?= https://github.com/spacemeshos/post-rs/releases/download/v$(POSTRS_SETUP_REV)/$(POSTRS_SERVICE_ZIP)
 
 ifeq ($(platform), windows)
 	POSTRS_SETUP_LIBS = post.h post.dll
+
 	POSTRS_PROFILER_BIN = profiler.exe
+	POSTRS_SERVICE_BIN = service.exe
 else
-	POSTRS_PROFILER_BIN = profiler
 	ifeq ($(platform), $(filter $(platform), macos macos-m1))
 		POSTRS_SETUP_LIBS = post.h libpost.dylib
 	else
 		POSTRS_SETUP_LIBS = post.h libpost.so
 	endif
+
+	POSTRS_PROFILER_BIN = profiler
+	POSTRS_SERVICE_BIN = service
 endif
 
 BINDIR_POSTRS_SETUP_LIBS = $(foreach X,$(POSTRS_SETUP_LIBS),$(BIN_DIR)$(X))
@@ -86,9 +92,16 @@ $(PROJ_DIR)$(POSTRS_SETUP_ZIP):
 $(BIN_DIR)$(POSTRS_PROFILER_BIN):
 	curl -sSL $(CURL_OPTIONS) $(POSTRS_PROFILER_URL) -o $(PROJ_DIR)$(POSTRS_PROFILER_ZIP)
 	unzip -o -j $(PROJ_DIR)$(POSTRS_PROFILER_ZIP) -d $(BIN_DIR)
+	chmod +x $(BIN_DIR)$(POSTRS_PROFILER_BIN)
+
+$(BIN_DIR)$(POSTRS_SERVICE_BIN):
+	curl -sSL $(CURL_OPTIONS) $(POSTRS_SERVICE_URL) -o $(PROJ_DIR)$(POSTRS_SERVICE_ZIP)
+	unzip -o -j $(PROJ_DIR)$(POSTRS_SERVICE_ZIP) -d $(BIN_DIR)
+	chmod +x $(BIN_DIR)$(POSTRS_SERVICE_BIN)
 
 get-postrs-lib: $(PROJ_DIR)$(POSTRS_SETUP_ZIP) $(BINDIR_POSTRS_SETUP_LIBS)
 get-postrs-profiler: $(BIN_DIR)$(POSTRS_PROFILER_BIN)
+get-postrs-service: $(BIN_DIR)$(POSTRS_SERVICE_BIN)
 
 go-env: get-postrs-lib
 	go env -w CGO_CFLAGS="$(CGO_CFLAGS)"

activation/interface.go

@@ -80,3 +80,7 @@ type SmeshingProvider interface {
 	SetCoinbase(coinbase types.Address)
 	UpdatePoETServers(ctx context.Context, endpoints []string) error
 }
+
+type PostClient interface {
+	Proof(ctx context.Context, challenge []byte) (*types.Post, *types.PostMetadata, error)
+}

activation/nipost_test.go

@@ -93,7 +93,6 @@ func TestNIPostBuilderWithMocks(t *testing.T) {
 }
 
 func TestPostSetup(t *testing.T) {
-	t.Parallel()
 	r := require.New(t)
 
 	postProvider := newTestPostManager(t)
@@ -139,8 +138,6 @@ func TestPostSetup(t *testing.T) {
 }
 
 func TestNIPostBuilderWithClients(t *testing.T) {
-	t.Parallel()
-
 	challenge := types.NIPostChallenge{
 		PublishEpoch: postGenesisEpoch + 2,
 	}
@@ -228,7 +225,6 @@ func TestNewNIPostBuilderNotInitialized(t *testing.T) {
 	if testing.Short() {
 		t.Skip()
 	}
-	t.Parallel()
 
 	r := require.New(t)
 

activation/post.go

@@ -12,11 +12,11 @@ import (
 	"github.com/spacemeshos/post/config"
 	"github.com/spacemeshos/post/initialization"
 	"github.com/spacemeshos/post/proving"
+	"go.uber.org/zap"
 
 	"github.com/spacemeshos/go-spacemesh/common/types"
 	"github.com/spacemeshos/go-spacemesh/datastore"
 	"github.com/spacemeshos/go-spacemesh/events"
-	"github.com/spacemeshos/go-spacemesh/log"
 	"github.com/spacemeshos/go-spacemesh/metrics/public"
 	"github.com/spacemeshos/go-spacemesh/sql"
 	"github.com/spacemeshos/go-spacemesh/sql/atxs"
@@ -94,7 +94,7 @@ type PostProviderID struct {
 }
 
 // String implements pflag.Value.String.
-func (id *PostProviderID) String() string {
+func (id PostProviderID) String() string {
 	if id.value == nil {
 		return ""
 	}
@@ -244,7 +244,7 @@ type PostSetupManager struct {
 	commitmentAtxId types.ATXID
 
 	cfg         PostConfig
-	logger      log.Log
+	logger      *zap.Logger
 	db          *datastore.CachedDB
 	goldenATXID types.ATXID
 
@@ -256,7 +256,7 @@ type PostSetupManager struct {
 }
 
 // NewPostSetupManager creates a new instance of PostSetupManager.
-func NewPostSetupManager(id types.NodeID, cfg PostConfig, logger log.Log, db *datastore.CachedDB, goldenATXID types.ATXID, provingOpts PostProvingOpts) (*PostSetupManager, error) {
+func NewPostSetupManager(id types.NodeID, cfg PostConfig, logger *zap.Logger, db *datastore.CachedDB, goldenATXID types.ATXID, provingOpts PostProvingOpts) (*PostSetupManager, error) {
 	mgr := &PostSetupManager{
 		id:          id,
 		cfg:         cfg,
@@ -358,13 +358,13 @@ func (mgr *PostSetupManager) StartSession(ctx context.Context) error {
 	if err != nil {
 		return err
 	}
-	mgr.logger.With().Info("post setup session starting",
-		log.String("node_id", mgr.id.String()),
-		log.String("commitment_atx", mgr.commitmentAtxId.String()),
-		log.String("data_dir", mgr.lastOpts.DataDir),
-		log.String("num_units", fmt.Sprintf("%d", mgr.lastOpts.NumUnits)),
-		log.String("labels_per_unit", fmt.Sprintf("%d", mgr.cfg.LabelsPerUnit)),
-		log.String("provider", fmt.Sprintf("%d", mgr.lastOpts.ProviderID)),
+	mgr.logger.Info("post setup session starting",
+		zap.Stringer("node_id", mgr.id),
+		zap.Stringer("commitment_atx", mgr.commitmentAtxId),
+		zap.String("data_dir", mgr.lastOpts.DataDir),
+		zap.Uint32("num_units", mgr.lastOpts.NumUnits),
+		zap.Uint64("labels_per_unit", mgr.cfg.LabelsPerUnit),
+		zap.Stringer("provider", mgr.lastOpts.ProviderID),
 	)
 	public.InitStart.Set(float64(mgr.lastOpts.NumUnits))
 	events.EmitInitStart(mgr.id, mgr.commitmentAtxId)
@@ -379,26 +379,26 @@ func (mgr *PostSetupManager) StartSession(ctx context.Context) error {
 		mgr.state = PostSetupStateStopped
 		return err
 	case errors.As(err, &errLabelMismatch):
-		mgr.logger.With().Error("post setup session failed due to an issue with the initialization provider", log.Err(errLabelMismatch))
+		mgr.logger.Error("post setup session failed due to an issue with the initialization provider", zap.Error(errLabelMismatch))
 		mgr.state = PostSetupStateError
 		events.EmitInitFailure(mgr.id, mgr.commitmentAtxId, errLabelMismatch)
 		return nil
 	case err != nil:
-		mgr.logger.With().Error("post setup session failed", log.Err(err))
+		mgr.logger.Error("post setup session failed", zap.Error(err))
 		mgr.state = PostSetupStateError
 		events.EmitInitFailure(mgr.id, mgr.commitmentAtxId, err)
 		return err
 	}
 	public.InitEnd.Set(float64(mgr.lastOpts.NumUnits))
 	events.EmitInitComplete()
 
-	mgr.logger.With().Info("post setup completed",
-		log.String("node_id", mgr.id.String()),
-		log.String("commitment_atx", mgr.commitmentAtxId.String()),
-		log.String("data_dir", mgr.lastOpts.DataDir),
-		log.String("num_units", fmt.Sprintf("%d", mgr.lastOpts.NumUnits)),
-		log.String("labels_per_unit", fmt.Sprintf("%d", mgr.cfg.LabelsPerUnit)),
-		log.String("provider", fmt.Sprintf("%d", mgr.lastOpts.ProviderID)),
+	mgr.logger.Info("post setup completed",
+		zap.Stringer("node_id", mgr.id),
+		zap.Stringer("commitment_atx", mgr.commitmentAtxId),
+		zap.String("data_dir", mgr.lastOpts.DataDir),
+		zap.Uint32("num_units", mgr.lastOpts.NumUnits),
+		zap.Uint64("labels_per_unit", mgr.cfg.LabelsPerUnit),
+		zap.Stringer("provider", mgr.lastOpts.ProviderID),
 	)
 	mgr.state = PostSetupStateComplete
 	return nil
@@ -420,15 +420,19 @@ func (mgr *PostSetupManager) PrepareInitializer(ctx context.Context, opts PostSe
 
 	// TODO(mafa): remove this, see https://github.com/spacemeshos/go-spacemesh/issues/4801
 	if opts.ProviderID.Value() != nil && *opts.ProviderID.Value() == -1 {
-		mgr.logger.Warning("DEPRECATED: auto-determining compute provider is deprecated, please specify a valid provider ID in the config file")
+		mgr.logger.Warn("DEPRECATED: auto-determining compute provider is deprecated, please specify a valid provider ID in the config file")
 
 		p, err := mgr.BestProvider()
 		if err != nil {
 			return err
 		}
 
-		mgr.logger.Warning("DEPRECATED: found best compute provider: id: %d, model: %v, device type: %v", p.ID, p.Model, p.DeviceType)
-		mgr.logger.Warning("DEPRECATED: please update your config file: {\"smeshing\": {\"smeshing-opts\": {\"smeshing-opts-provider\": %d }}}", p.ID)
+		mgr.logger.Warn("DEPRECATED: found best compute provider",
+			zap.Uint32("id", p.ID),
+			zap.String("model", p.Model),
+			zap.Stringer("device type", p.DeviceType),
+		)
+		mgr.logger.Sugar().Warnf("DEPRECATED: please update your config file: {\"smeshing\": {\"smeshing-opts\": {\"smeshing-opts-provider\": %d }}}", p.ID)
 		opts.ProviderID.SetInt64(int64(p.ID))
 	}
 
@@ -443,7 +447,7 @@ func (mgr *PostSetupManager) PrepareInitializer(ctx context.Context, opts PostSe
 		initialization.WithCommitmentAtxId(mgr.commitmentAtxId.Bytes()),
 		initialization.WithConfig(mgr.cfg.ToConfig()),
 		initialization.WithInitOpts(opts.ToInitOpts()),
-		initialization.WithLogger(mgr.logger.Zap()),
+		initialization.WithLogger(mgr.logger),
 	)
 	if err != nil {
 		mgr.state = PostSetupStateError
@@ -499,7 +503,7 @@ func (mgr *PostSetupManager) findCommitmentAtx(ctx context.Context) (types.ATXID
 	atx, err := atxs.GetIDWithMaxHeight(mgr.db, types.EmptyNodeID)
 	switch {
 	case errors.Is(err, sql.ErrNotFound):
-		mgr.logger.With().Info("using golden atx as commitment atx")
+		mgr.logger.Info("using golden atx as commitment atx")
 		return mgr.goldenATXID, nil
 	case err != nil:
 		return types.EmptyATXID, fmt.Errorf("get commitment atx: %w", err)
@@ -540,7 +544,7 @@ func (mgr *PostSetupManager) GenerateProof(ctx context.Context, challenge []byte
 	}
 	opts = append(opts, options...)
 
-	proof, proofMetadata, err := proving.Generate(ctx, challenge, mgr.cfg.ToConfig(), mgr.logger.Zap(), opts...)
+	proof, proofMetadata, err := proving.Generate(ctx, challenge, mgr.cfg.ToConfig(), mgr.logger, opts...)
 	if err != nil {
 		return nil, nil, fmt.Errorf("generate proof: %w", err)
 	}