Files
LocalAI/tests/e2e/distributed/backend_logs_test.go
T
Ettore Di Giacinto a0c7b2709b test(distributed): scope the log-subscriber wait and mark the race it works around
Three corrections from review of the previous commit.

The lock-order comment on SubscriberCount claimed no path takes s.mu and a
buffer lock together. Subscribe does exactly that, holding s.mu.RLock across
replica registrations that take buf.mu. State the rule that is actually true —
s.mu precedes any buffer lock, so counting after releasing it preserves the
order — and say what follows from it: the total is a sample, not a snapshot.

waitForLogSubscriber read as general-purpose but unblocks on the first
registered subscription. Subscribe attaches the exact-key buffer and each
replica buffer one at a time, so for a replicated model the count goes positive
while later replicas are still unattached and the race survives. Rename it
waitForSingleLogSubscriber, document that it holds only where Subscribe
resolves to one buffer, and assert on exactly 1: misuse then fails loudly on
the count rather than going quietly back to being flaky. Taking the expected
count as a parameter was the alternative, but that makes callers predict a
store-internal number and an under-count fails the same silent way as the
original bug.

The snapshot-then-subscribe race had no artifact outside a report, and review
found a second site carrying it. Mark both handlers identically, including the
point that swapping the two calls duplicates rather than drops and so is not
the fix. The race itself is left alone; this branch stays test infrastructure.

Assisted-by: Claude Opus 5 [claude-code]
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2026-09-20 03:05:33 +00:00

566 lines
19 KiB
Go

package distributed_test
import (
"context"
"encoding/json"
"fmt"
"net"
"net/http"
"net/http/httptest"
"net/url"
"os"
"time"
"github.com/gorilla/websocket"
"github.com/labstack/echo/v4"
"github.com/mudler/LocalAI/core/http/endpoints/localai"
"github.com/mudler/LocalAI/core/services/nodes"
"github.com/mudler/LocalAI/pkg/model"
. "github.com/onsi/ginkgo/v2"
. "github.com/onsi/gomega"
pgdriver "gorm.io/driver/postgres"
"gorm.io/gorm"
"gorm.io/gorm/logger"
)
// waitForSingleLogSubscriber blocks until the worker's WebSocket log handler has
// registered its subscription on the store.
//
// The handler writes the "initial" batch first and subscribes only afterwards,
// so a line appended the instant that batch lands is buffered but never
// streamed, and the spec then waits out its full read deadline. Measured at
// roughly one run in seventeen with `--repeat`, which is far too often for CI.
// Waiting on the subscription removes the race from the spec; the handler's own
// snapshot/subscribe window is a separate production question, marked at both
// production sites.
//
// Only valid where BackendLogStore.Subscribe resolves modelID to exactly ONE
// buffer: a bare model ID with no "<modelID>#N" replica buffers in the store, or
// a full process key. Subscribe registers the exact-key buffer and each replica
// buffer one at a time, so for a model that does have replicas the count goes
// positive while later replicas are still unattached and the race survives.
// Hence the assertion is on exactly 1 rather than "at least 1": a spec that
// misapplies this to a replicated model fails loudly on the count instead of
// going quietly back to being flaky.
func waitForSingleLogSubscriber(logStore *model.BackendLogStore, modelID string) {
GinkgoHelper()
Eventually(func() int { return logStore.SubscriberCount(modelID) }, "10s", "5ms").
Should(Equal(1), "the WebSocket handler never subscribed to %q exactly once", modelID)
}
var _ = Describe("Distributed Backend Log Streaming", Label("Distributed"), func() {
Context("Worker HTTP log endpoints", func() {
var (
logStore *model.BackendLogStore
serverAddr string
cleanup func()
token string
)
BeforeEach(func() {
token = "test-secret-token"
logStore = model.NewBackendLogStore(1000)
// Populate test log lines
logStore.AppendLine("model-a", "stdout", "loading model...")
logStore.AppendLine("model-a", "stderr", "warning: something")
logStore.AppendLine("model-a", "stdout", "model loaded successfully")
logStore.AppendLine("model-b", "stdout", "hello from model-b")
var err error
serverAddr, cleanup, err = startTestFileTransferServerWithLogs(token, logStore)
Expect(err).ToNot(HaveOccurred())
})
AfterEach(func() {
if cleanup != nil {
cleanup()
}
})
It("should list models with logs via GET /v1/backend-logs", func() {
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs", serverAddr), nil)
Expect(err).ToNot(HaveOccurred())
req.Header.Set("Authorization", "Bearer "+token)
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusOK))
var models []string
Expect(json.NewDecoder(resp.Body).Decode(&models)).To(Succeed())
Expect(models).To(ContainElement("model-a"))
Expect(models).To(ContainElement("model-b"))
})
It("should return log lines for a model via GET /v1/backend-logs/{modelId}", func() {
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs/model-a", serverAddr), nil)
Expect(err).ToNot(HaveOccurred())
req.Header.Set("Authorization", "Bearer "+token)
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusOK))
var lines []model.BackendLogLine
Expect(json.NewDecoder(resp.Body).Decode(&lines)).To(Succeed())
Expect(lines).To(HaveLen(3))
Expect(lines[0].Stream).To(Equal("stdout"))
Expect(lines[0].Text).To(Equal("loading model..."))
Expect(lines[1].Stream).To(Equal("stderr"))
Expect(lines[1].Text).To(Equal("warning: something"))
Expect(lines[2].Stream).To(Equal("stdout"))
Expect(lines[2].Text).To(Equal("model loaded successfully"))
})
It("should return empty array for unknown model", func() {
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs/nonexistent", serverAddr), nil)
Expect(err).ToNot(HaveOccurred())
req.Header.Set("Authorization", "Bearer "+token)
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusOK))
var lines []model.BackendLogLine
Expect(json.NewDecoder(resp.Body).Decode(&lines)).To(Succeed())
Expect(lines).To(BeEmpty())
})
It("should reject requests without bearer token", func() {
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs", serverAddr), nil)
Expect(err).ToNot(HaveOccurred())
// No Authorization header
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusUnauthorized))
})
It("should reject requests with wrong bearer token", func() {
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs", serverAddr), nil)
Expect(err).ToNot(HaveOccurred())
req.Header.Set("Authorization", "Bearer wrong-token")
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusUnauthorized))
})
It("should handle URL-encoded model IDs", func() {
// Add a model with special characters in the name
logStore.AppendLine("my-org/model:latest", "stdout", "special model log")
encodedModelID := url.PathEscape("my-org/model:latest")
req, err := http.NewRequest("GET", fmt.Sprintf("http://%s/v1/backend-logs/%s", serverAddr, encodedModelID), nil)
Expect(err).ToNot(HaveOccurred())
req.Header.Set("Authorization", "Bearer "+token)
resp, err := http.DefaultClient.Do(req)
Expect(err).ToNot(HaveOccurred())
defer resp.Body.Close()
Expect(resp.StatusCode).To(Equal(http.StatusOK))
var lines []model.BackendLogLine
Expect(json.NewDecoder(resp.Body).Decode(&lines)).To(Succeed())
Expect(lines).To(HaveLen(1))
Expect(lines[0].Text).To(Equal("special model log"))
})
})
Context("Worker WebSocket log streaming", func() {
var (
logStore *model.BackendLogStore
serverAddr string
cleanup func()
token string
)
BeforeEach(func() {
token = "test-ws-token"
logStore = model.NewBackendLogStore(1000)
// Pre-populate some lines
logStore.AppendLine("ws-model", "stdout", "line-1")
logStore.AppendLine("ws-model", "stderr", "line-2")
var err error
serverAddr, cleanup, err = startTestFileTransferServerWithLogs(token, logStore)
Expect(err).ToNot(HaveOccurred())
})
AfterEach(func() {
if cleanup != nil {
cleanup()
}
})
It("should stream initial lines and new lines via WebSocket", func() {
wsURL := fmt.Sprintf("ws://%s/v1/backend-logs/ws-model/ws", serverAddr)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
headers := http.Header{}
headers.Set("Authorization", "Bearer "+token)
conn, resp, err := dialer.Dial(wsURL, headers)
Expect(err).ToNot(HaveOccurred())
Expect(resp.StatusCode).To(Equal(http.StatusSwitchingProtocols))
defer conn.Close()
// Read initial message
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var initialMsg map[string]json.RawMessage
err = conn.ReadJSON(&initialMsg)
Expect(err).ToNot(HaveOccurred())
var msgType string
Expect(json.Unmarshal(initialMsg["type"], &msgType)).To(Succeed())
Expect(msgType).To(Equal("initial"))
var initialLines []model.BackendLogLine
Expect(json.Unmarshal(initialMsg["lines"], &initialLines)).To(Succeed())
Expect(initialLines).To(HaveLen(2))
Expect(initialLines[0].Text).To(Equal("line-1"))
Expect(initialLines[1].Text).To(Equal("line-2"))
// Now append a new line and verify it arrives via WebSocket
waitForSingleLogSubscriber(logStore, "ws-model")
logStore.AppendLine("ws-model", "stdout", "line-3-realtime")
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var lineMsg map[string]json.RawMessage
err = conn.ReadJSON(&lineMsg)
Expect(err).ToNot(HaveOccurred())
Expect(json.Unmarshal(lineMsg["type"], &msgType)).To(Succeed())
Expect(msgType).To(Equal("line"))
var streamedLine model.BackendLogLine
Expect(json.Unmarshal(lineMsg["line"], &streamedLine)).To(Succeed())
Expect(streamedLine.Text).To(Equal("line-3-realtime"))
Expect(streamedLine.Stream).To(Equal("stdout"))
})
It("should reject WebSocket connection without token", func() {
wsURL := fmt.Sprintf("ws://%s/v1/backend-logs/ws-model/ws", serverAddr)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
_, resp, err := dialer.Dial(wsURL, nil)
Expect(err).To(HaveOccurred())
if resp != nil {
Expect(resp.StatusCode).To(Equal(http.StatusUnauthorized))
}
})
It("should handle client disconnect gracefully", func() {
wsURL := fmt.Sprintf("ws://%s/v1/backend-logs/ws-model/ws", serverAddr)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
headers := http.Header{}
headers.Set("Authorization", "Bearer "+token)
conn, _, err := dialer.Dial(wsURL, headers)
Expect(err).ToNot(HaveOccurred())
// Read initial message
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var initialMsg map[string]json.RawMessage
Expect(conn.ReadJSON(&initialMsg)).To(Succeed())
// Close connection abruptly
conn.Close()
// Append more lines — should not panic
logStore.AppendLine("ws-model", "stdout", "after-disconnect")
time.Sleep(100 * time.Millisecond)
// If we got here without panic, the test passes
})
It("should stream lines only for the requested model", func() {
logStore.AppendLine("other-model", "stdout", "other model log")
wsURL := fmt.Sprintf("ws://%s/v1/backend-logs/ws-model/ws", serverAddr)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
headers := http.Header{}
headers.Set("Authorization", "Bearer "+token)
conn, _, err := dialer.Dial(wsURL, headers)
Expect(err).ToNot(HaveOccurred())
defer conn.Close()
// Read initial message
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var initialMsg map[string]json.RawMessage
Expect(conn.ReadJSON(&initialMsg)).To(Succeed())
// Append line to a different model
waitForSingleLogSubscriber(logStore, "ws-model")
logStore.AppendLine("other-model", "stdout", "should not appear")
// Append line to our model
logStore.AppendLine("ws-model", "stdout", "should appear")
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var lineMsg map[string]json.RawMessage
Expect(conn.ReadJSON(&lineMsg)).To(Succeed())
var streamedLine model.BackendLogLine
Expect(json.Unmarshal(lineMsg["line"], &streamedLine)).To(Succeed())
Expect(streamedLine.Text).To(Equal("should appear"))
})
})
Context("Frontend proxy REST endpoints", func() {
var (
pgInfra *TestInfra
db *gorm.DB
registry *nodes.NodeRegistry
logStore *model.BackendLogStore
workerAddr string
workerClean func()
token string
)
BeforeEach(func() {
pgInfra = SetupInfra("localai_backend_logs_test")
var err error
db, err = gorm.Open(pgdriver.Open(pgInfra.PGURL), &gorm.Config{
Logger: logger.Default.LogMode(logger.Silent),
})
Expect(err).ToNot(HaveOccurred())
registry, err = nodes.NewNodeRegistry(db)
Expect(err).ToNot(HaveOccurred())
token = "proxy-test-token"
logStore = model.NewBackendLogStore(1000)
logStore.AppendLine("remote-model", "stdout", "remote log line 1")
logStore.AppendLine("remote-model", "stderr", "remote log line 2")
workerAddr, workerClean, err = startTestFileTransferServerWithLogs(token, logStore)
Expect(err).ToNot(HaveOccurred())
})
AfterEach(func() {
if workerClean != nil {
workerClean()
}
})
It("should proxy backend-logs list from worker via node ID", func() {
// Register a node with HTTPAddress pointing to our test worker server
node := &nodes.BackendNode{
Name: "log-test-node",
Address: "127.0.0.1:50051", // gRPC address (unused here)
HTTPAddress: workerAddr,
}
Expect(registry.Register(context.Background(), node, true)).To(Succeed())
// Create an Echo test server with the proxy endpoint
e := echo.New()
e.GET("/api/nodes/:id/backend-logs", localai.NodeBackendLogsListEndpoint(registry, token))
req := httptest.NewRequest("GET", fmt.Sprintf("/api/nodes/%s/backend-logs", node.ID), nil)
rec := httptest.NewRecorder()
e.ServeHTTP(rec, req)
Expect(rec.Code).To(Equal(http.StatusOK))
var models []string
Expect(json.NewDecoder(rec.Body).Decode(&models)).To(Succeed())
Expect(models).To(ContainElement("remote-model"))
})
It("should proxy backend-logs lines from worker via node ID", func() {
node := &nodes.BackendNode{
Name: "log-lines-node",
Address: "127.0.0.1:50051",
HTTPAddress: workerAddr,
}
Expect(registry.Register(context.Background(), node, true)).To(Succeed())
e := echo.New()
e.GET("/api/nodes/:id/backend-logs/:modelId", localai.NodeBackendLogsLinesEndpoint(registry, token))
req := httptest.NewRequest("GET", fmt.Sprintf("/api/nodes/%s/backend-logs/remote-model", node.ID), nil)
rec := httptest.NewRecorder()
e.ServeHTTP(rec, req)
Expect(rec.Code).To(Equal(http.StatusOK))
var lines []model.BackendLogLine
Expect(json.NewDecoder(rec.Body).Decode(&lines)).To(Succeed())
Expect(lines).To(HaveLen(2))
Expect(lines[0].Text).To(Equal("remote log line 1"))
Expect(lines[1].Text).To(Equal("remote log line 2"))
})
It("should return 404 for unknown node ID", func() {
e := echo.New()
e.GET("/api/nodes/:id/backend-logs", localai.NodeBackendLogsListEndpoint(registry, token))
req := httptest.NewRequest("GET", "/api/nodes/nonexistent-id/backend-logs", nil)
rec := httptest.NewRecorder()
e.ServeHTTP(rec, req)
Expect(rec.Code).To(Equal(http.StatusNotFound))
})
})
Context("Frontend WebSocket proxy (end-to-end)", func() {
var (
wsInfra *TestInfra
db *gorm.DB
registry *nodes.NodeRegistry
logStore *model.BackendLogStore
workerAddr string
workerClean func()
token string
echoServer *http.Server
echoAddr string
)
BeforeEach(func() {
wsInfra = SetupInfra("localai_ws_proxy_test")
var err error
db, err = gorm.Open(pgdriver.Open(wsInfra.PGURL), &gorm.Config{
Logger: logger.Default.LogMode(logger.Silent),
})
Expect(err).ToNot(HaveOccurred())
registry, err = nodes.NewNodeRegistry(db)
Expect(err).ToNot(HaveOccurred())
token = "ws-proxy-token"
logStore = model.NewBackendLogStore(1000)
logStore.AppendLine("proxy-model", "stdout", "initial line from worker")
workerAddr, workerClean, err = startTestFileTransferServerWithLogs(token, logStore)
Expect(err).ToNot(HaveOccurred())
// Start Echo server with the WebSocket proxy route
e := echo.New()
e.GET("/ws/nodes/:id/backend-logs/:modelId", localai.NodeBackendLogsWSEndpoint(registry, token))
lis, err := net.Listen("tcp", "127.0.0.1:0")
Expect(err).ToNot(HaveOccurred())
echoAddr = lis.Addr().String()
echoServer = &http.Server{Handler: e}
go echoServer.Serve(lis)
})
AfterEach(func() {
if echoServer != nil {
echoServer.Close()
}
if workerClean != nil {
workerClean()
}
})
It("should proxy WebSocket log stream from worker through frontend", func() {
// Register node
node := &nodes.BackendNode{
Name: "ws-proxy-node",
Address: "127.0.0.1:50051",
HTTPAddress: workerAddr,
}
Expect(registry.Register(context.Background(), node, true)).To(Succeed())
// Connect WebSocket to the frontend proxy
wsURL := fmt.Sprintf("ws://%s/ws/nodes/%s/backend-logs/proxy-model", echoAddr, node.ID)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
conn, _, err := dialer.Dial(wsURL, nil)
Expect(err).ToNot(HaveOccurred())
defer conn.Close()
// Read initial message (proxied from worker)
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var initialMsg map[string]json.RawMessage
Expect(conn.ReadJSON(&initialMsg)).To(Succeed())
var msgType string
Expect(json.Unmarshal(initialMsg["type"], &msgType)).To(Succeed())
Expect(msgType).To(Equal("initial"))
var initialLines []model.BackendLogLine
Expect(json.Unmarshal(initialMsg["lines"], &initialLines)).To(Succeed())
Expect(initialLines).To(HaveLen(1))
Expect(initialLines[0].Text).To(Equal("initial line from worker"))
// Append a new line on the worker's log store
waitForSingleLogSubscriber(logStore, "proxy-model")
logStore.AppendLine("proxy-model", "stderr", "realtime via proxy")
// Read the streamed line through the proxy
conn.SetReadDeadline(time.Now().Add(5 * time.Second))
var lineMsg map[string]json.RawMessage
Expect(conn.ReadJSON(&lineMsg)).To(Succeed())
Expect(json.Unmarshal(lineMsg["type"], &msgType)).To(Succeed())
Expect(msgType).To(Equal("line"))
var streamedLine model.BackendLogLine
Expect(json.Unmarshal(lineMsg["line"], &streamedLine)).To(Succeed())
Expect(streamedLine.Text).To(Equal("realtime via proxy"))
Expect(streamedLine.Stream).To(Equal("stderr"))
})
It("should return error for unknown node in WebSocket proxy", func() {
wsURL := fmt.Sprintf("ws://%s/ws/nodes/nonexistent-node/backend-logs/some-model", echoAddr)
dialer := websocket.Dialer{HandshakeTimeout: 5 * time.Second}
_, resp, err := dialer.Dial(wsURL, nil)
Expect(err).To(HaveOccurred())
if resp != nil {
// Should get a non-101 status (404 or similar)
Expect(resp.StatusCode).ToNot(Equal(http.StatusSwitchingProtocols))
}
})
})
})
// startTestFileTransferServerWithLogs starts the real nodes.StartFileTransferServerWithListener
// with a BackendLogStore, using a temporary staging directory.
// Returns the server address, cleanup function, and error.
func startTestFileTransferServerWithLogs(token string, logStore *model.BackendLogStore) (string, func(), error) {
stagingDir, err := os.MkdirTemp("", "logs-test-staging-*")
if err != nil {
return "", nil, err
}
// Listen on a free port and pass the listener directly to avoid TOCTOU race.
lis, err := net.Listen("tcp", "127.0.0.1:0")
if err != nil {
os.RemoveAll(stagingDir)
return "", nil, err
}
addr := lis.Addr().String()
server, err := nodes.StartFileTransferServerWithListener(lis, stagingDir, stagingDir, stagingDir, token, 0, logStore)
if err != nil {
os.RemoveAll(stagingDir)
return "", nil, err
}
cleanup := func() {
nodes.ShutdownFileTransferServer(server)
os.RemoveAll(stagingDir)
}
return addr, cleanup, nil
}