feat(openflare): migrate of_node_access_logs to ClickHouse

Move node access log storage from PostgreSQL/SQLite to ClickHouse
(database: openflare). System startup now requires a healthy ClickHouse
connection and auto-initializes the schema. Agent heartbeat writes access
logs via batch insert; goose migration drops the legacy relational table.
This commit is contained in:
ryan
2026-06-19 11:45:48 +08:00
parent d78449cbc9
commit 425ca89765
24 changed files with 1224 additions and 458 deletions
@@ -16,7 +16,6 @@ import (
"time"
"github.com/Rain-kl/Wavelet/internal/apps/admin"
"github.com/Rain-kl/Wavelet/internal/config"
"github.com/Rain-kl/Wavelet/internal/db"
"github.com/Rain-kl/Wavelet/internal/model"
"github.com/Rain-kl/Wavelet/pkg/logger"
@@ -288,9 +287,8 @@ func fetchAccessLogDetails(ctx context.Context, whereClause string, args []inter
// @Failure 403 {object} response.Any "无管理员权限"
// @Router /api/v1/admin/logs/access [get]
func GetAccessLogs(c *gin.Context) {
// 1. 检查 ClickHouse 是否启用
if !config.Config.ClickHouse.Enabled || db.ChConn == nil {
response.AbortWithError(c, http.StatusBadRequest, "ClickHouse 存储服务未启用,无法检索访问日志")
if db.ChConn == nil {
response.AbortWithError(c, http.StatusInternalServerError, "ClickHouse 未初始化,无法检索访问日志")
return
}
@@ -388,9 +386,8 @@ type logsAnalyticsResponse struct {
// @Failure 403 {object} response.Any "无管理员权限"
// @Router /api/v1/admin/logs/analytics [get]
func GetLogsAnalytics(c *gin.Context) {
// 1. 检查 ClickHouse 是否启用
if !config.Config.ClickHouse.Enabled || db.ChConn == nil {
response.AbortWithError(c, http.StatusBadRequest, "ClickHouse 存储服务未启用,无法获取分析数据")
if db.ChConn == nil {
response.AbortWithError(c, http.StatusInternalServerError, "ClickHouse 未初始化,无法获取分析数据")
return
}
@@ -124,6 +124,12 @@ func PersistHeartbeatObservability(ctx context.Context, nodeID string, payload N
return
}
accessLogRecords, err := buildNodeAccessLogRecords(nodeID, payload.AccessLogs, payload.BufferedObservability, reportedAt)
if err != nil {
zap.L().Error("build heartbeat access logs failed", zap.String("node_id", nodeID), zap.Error(err))
return
}
if err := conn.Transaction(func(tx *gorm.DB) error {
if err := persistNodeSystemProfile(tx, nodeID, payload.Profile, reportedAt); err != nil {
return err
@@ -140,9 +146,6 @@ func PersistHeartbeatObservability(ctx context.Context, nodeID string, payload N
if err := persistNodeTrafficReport(tx, nodeID, payload.TrafficReport, reportedAt); err != nil {
return err
}
if err := persistNodeAccessLogs(tx, nodeID, payload.AccessLogs, reportedAt); err != nil {
return err
}
if payload.HealthEvents != nil {
if err := reconcileNodeHealthEvents(tx, nodeID, payload.HealthEvents, reportedAt); err != nil {
return err
@@ -151,6 +154,11 @@ func PersistHeartbeatObservability(ctx context.Context, nodeID string, payload N
return nil
}); err != nil {
zap.L().Error("persist heartbeat observability failed", zap.String("node_id", nodeID), zap.Error(err))
return
}
if err := persistNodeAccessLogs(ctx, nodeID, accessLogRecords, reportedAt); err != nil {
zap.L().Error("persist heartbeat access logs failed", zap.String("node_id", nodeID), zap.Error(err))
}
}
@@ -165,9 +173,7 @@ func persistBufferedObservability(tx *gorm.DB, nodeID string, records []Buffered
if err := persistNodeTrafficReport(tx, nodeID, record.TrafficReport, reportedAt); err != nil {
return err
}
if err := persistNodeAccessLogs(tx, nodeID, record.AccessLogs, reportedAt); err != nil {
return err
}
}
return nil
}
@@ -278,10 +284,15 @@ func persistNodeTrafficReport(tx *gorm.DB, nodeID string, report *NodeTrafficRep
return tx.Create(record).Error
}
func persistNodeAccessLogs(tx *gorm.DB, nodeID string, logs []NodeAccessLog, reportedAt time.Time) error {
if len(logs) == 0 {
return nil
func buildNodeAccessLogRecords(nodeID string, direct []NodeAccessLog, buffered []BufferedObservabilityRecord, reportedAt time.Time) ([]*model.OpenFlareAccessLog, error) {
total := len(direct)
for _, record := range buffered {
total += len(record.AccessLogs)
}
if total == 0 {
return nil, nil
}
resolver, err := newAccessLogRegionResolver()
if err != nil {
slog.Warn("initialize access log geo resolver failed", "node_id", nodeID, "error", err)
@@ -289,31 +300,40 @@ func persistNodeAccessLogs(tx *gorm.DB, nodeID string, logs []NodeAccessLog, rep
if resolver != nil {
defer resolver.Close()
}
for _, item := range logs {
record := &model.OpenFlareAccessLog{
NodeID: nodeID,
LoggedAt: timeFromUnix(item.LoggedAtUnix, reportedAt),
RemoteAddr: strings.TrimSpace(item.RemoteAddr),
Region: "",
Host: strings.TrimSpace(item.Host),
Path: truncateForDatabase(strings.TrimSpace(item.Path), accessLogPathMaxLength),
StatusCode: item.StatusCode,
}
if resolver != nil {
record.Region = resolver.Resolve(record.RemoteAddr)
}
exists, err := accessLogExists(tx, record)
if err != nil {
return err
}
if exists {
continue
}
if err := tx.Create(record).Error; err != nil {
return err
records := make([]*model.OpenFlareAccessLog, 0, total)
appendLogs := func(logs []NodeAccessLog) {
for _, item := range logs {
record := &model.OpenFlareAccessLog{
NodeID: nodeID,
LoggedAt: timeFromUnix(item.LoggedAtUnix, reportedAt),
RemoteAddr: strings.TrimSpace(item.RemoteAddr),
Region: "",
Host: strings.TrimSpace(item.Host),
Path: truncateForDatabase(strings.TrimSpace(item.Path), accessLogPathMaxLength),
StatusCode: item.StatusCode,
}
if resolver != nil {
record.Region = resolver.Resolve(record.RemoteAddr)
}
records = append(records, record)
}
}
_, err = deleteAccessLogsByNodeBefore(tx, nodeID, reportedAt.Add(-nodeAccessLogRetentionWindow))
appendLogs(direct)
for _, record := range buffered {
appendLogs(record.AccessLogs)
}
return records, nil
}
func persistNodeAccessLogs(ctx context.Context, nodeID string, records []*model.OpenFlareAccessLog, reportedAt time.Time) error {
if len(records) == 0 {
return nil
}
if err := model.InsertOpenFlareAccessLogsBatch(ctx, records); err != nil {
return err
}
_, err := model.DeleteOpenFlareAccessLogsByNodeBefore(ctx, nodeID, reportedAt.Add(-nodeAccessLogRetentionWindow))
return err
}
@@ -434,30 +454,6 @@ func requestReportExists(tx *gorm.DB, nodeID string, windowStartedAt, windowEnde
return count > 0, nil
}
func deleteAccessLogsByNodeBefore(tx *gorm.DB, nodeID string, before time.Time) (int64, error) {
result := tx.Where("node_id = ? AND logged_at < ?", nodeID, before).Delete(&model.OpenFlareAccessLog{})
return result.RowsAffected, result.Error
}
func accessLogExists(tx *gorm.DB, record *model.OpenFlareAccessLog) (bool, error) {
var count int64
if err := tx.Model(&model.OpenFlareAccessLog{}).
Where(
"node_id = ? AND logged_at = ? AND remote_addr = ? AND host = ? AND path = ? AND status_code = ?",
record.NodeID,
record.LoggedAt,
record.RemoteAddr,
record.Host,
record.Path,
record.StatusCode,
).
Limit(1).
Count(&count).Error; err != nil {
return false, err
}
return count > 0, nil
}
func normalizeHealthEventType(eventType string) string {
eventType = strings.TrimSpace(strings.ToLower(eventType))
eventType = strings.ReplaceAll(eventType, " ", "_")
@@ -34,21 +34,22 @@ func TestDatabaseAutoCleanupHandlerDeletesRowsWhenEnabled(t *testing.T) {
DisableForeignKeyConstraintWhenMigrating: true,
})
require.NoError(t, err)
require.NoError(t, sqliteDB.AutoMigrate(&model.OpenFlareAccessLog{}))
db.SetDB(sqliteDB)
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
t.Cleanup(func() {
resetAccessLogStore()
db.SetDB(nil)
})
now := time.Now().UTC()
require.NoError(t, db.DB(context.Background()).Create(&model.OpenFlareAccessLog{
require.NoError(t, model.InsertOpenFlareAccessLogsBatch(context.Background(), []*model.OpenFlareAccessLog{{
NodeID: "node-a",
LoggedAt: now.Add(-48 * time.Hour),
RemoteAddr: "203.0.113.10",
Host: "example.com",
Path: "/access",
StatusCode: 200,
}).Error)
}}))
previousEnabled := model.DatabaseAutoCleanupEnabled
previousRetentionDays := model.DatabaseAutoCleanupRetentionDays
@@ -24,7 +24,9 @@ func setupDashboardTestDB(t *testing.T) func() {
require.NoError(t, sqliteDB.AutoMigrate(&model.OpenFlareNode{}))
db.SetDB(sqliteDB)
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
return func() {
resetAccessLogStore()
db.SetDB(nil)
}
}
@@ -35,8 +35,10 @@ func setupNodeTestDB(t *testing.T) func() {
db.SetDB(sqliteDB)
option.ResetInitializationForTest()
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
return func() {
resetAccessLogStore()
db.SetDB(nil)
option.ResetInitializationForTest()
}
@@ -101,26 +101,28 @@ func TestCleanupDatabaseObservabilityDeletesRows(t *testing.T) {
defer cleanup()
ctx := context.Background()
sqliteDB := db.DB(ctx)
require.NoError(t, sqliteDB.AutoMigrate(&model.OpenFlareAccessLog{}))
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
defer resetAccessLogStore()
now := time.Now().UTC()
require.NoError(t, sqliteDB.Create(&model.OpenFlareAccessLog{
NodeID: "node-a",
LoggedAt: now.Add(-10 * 24 * time.Hour),
RemoteAddr: "203.0.113.1",
Host: "example.com",
Path: "/old",
StatusCode: 200,
}).Error)
require.NoError(t, sqliteDB.Create(&model.OpenFlareAccessLog{
NodeID: "node-a",
LoggedAt: now.Add(-2 * time.Hour),
RemoteAddr: "203.0.113.2",
Host: "example.com",
Path: "/recent",
StatusCode: 200,
}).Error)
require.NoError(t, model.InsertOpenFlareAccessLogsBatch(ctx, []*model.OpenFlareAccessLog{
{
NodeID: "node-a",
LoggedAt: now.Add(-10 * 24 * time.Hour),
RemoteAddr: "203.0.113.1",
Host: "example.com",
Path: "/old",
StatusCode: 200,
},
{
NodeID: "node-a",
LoggedAt: now.Add(-2 * time.Hour),
RemoteAddr: "203.0.113.2",
Host: "example.com",
Path: "/recent",
StatusCode: 200,
},
}))
retention := 7
result, err := cleanupDatabaseObservability(ctx, databaseCleanupInput{
@@ -24,12 +24,13 @@ func setupDatabaseCleanupTestDB(t *testing.T) context.Context {
})
require.NoError(t, err)
require.NoError(t, sqliteDB.AutoMigrate(
&model.OpenFlareAccessLog{},
&model.OpenFlareMetricSnapshot{},
&model.OpenFlareRequestReport{},
))
db.SetDB(sqliteDB)
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
t.Cleanup(func() {
resetAccessLogStore()
db.SetDB(nil)
})
return context.Background()
@@ -69,22 +70,24 @@ func TestCleanupDatabaseObservabilityDeletesAllRowsWhenRetentionMissing(t *testi
ctx := setupDatabaseCleanupTestDB(t)
now := time.Now().UTC()
require.NoError(t, db.DB(ctx).Create(&model.OpenFlareAccessLog{
NodeID: "node-a",
LoggedAt: now.Add(-3 * time.Hour),
RemoteAddr: "203.0.113.1",
Host: "example.com",
Path: "/one",
StatusCode: 200,
}).Error)
require.NoError(t, db.DB(ctx).Create(&model.OpenFlareAccessLog{
NodeID: "node-a",
LoggedAt: now.Add(-2 * time.Hour),
RemoteAddr: "203.0.113.2",
Host: "example.com",
Path: "/two",
StatusCode: 502,
}).Error)
require.NoError(t, model.InsertOpenFlareAccessLogsBatch(ctx, []*model.OpenFlareAccessLog{
{
NodeID: "node-a",
LoggedAt: now.Add(-3 * time.Hour),
RemoteAddr: "203.0.113.1",
Host: "example.com",
Path: "/one",
StatusCode: 200,
},
{
NodeID: "node-a",
LoggedAt: now.Add(-2 * time.Hour),
RemoteAddr: "203.0.113.2",
Host: "example.com",
Path: "/two",
StatusCode: 502,
},
}))
result, err := CleanupDatabaseObservability(ctx, DatabaseCleanupInput{
Target: DatabaseCleanupTargetAccessLogs,
@@ -102,14 +105,14 @@ func TestRunDatabaseAutoCleanupOnceDeletesAllObservabilityTargets(t *testing.T)
ctx := setupDatabaseCleanupTestDB(t)
now := time.Now().UTC()
require.NoError(t, db.DB(ctx).Create(&model.OpenFlareAccessLog{
require.NoError(t, model.InsertOpenFlareAccessLogsBatch(ctx, []*model.OpenFlareAccessLog{{
NodeID: "node-a",
LoggedAt: now.Add(-48 * time.Hour),
RemoteAddr: "203.0.113.10",
Host: "example.com",
Path: "/access",
StatusCode: 200,
}).Error)
}}))
require.NoError(t, db.DB(ctx).Create(&model.OpenFlareMetricSnapshot{
NodeID: "node-a",
CapturedAt: now.Add(-48 * time.Hour),
@@ -29,11 +29,12 @@ func setupIPGroupSyncTestDB(t *testing.T) func() {
require.NoError(t, sqliteDB.AutoMigrate(
&model.OpenFlareWAFRuleGroup{},
&model.OpenFlareWAFIPGroup{},
&model.OpenFlareAccessLog{},
))
db.SetDB(sqliteDB)
resetAccessLogStore := model.SetAccessLogStoreForTest(model.NewMemoryAccessLogStore())
return func() {
resetAccessLogStore()
db.SetDB(nil)
}
}
@@ -191,18 +192,20 @@ func TestListDueOpenFlareWAFIPGroups(t *testing.T) {
func seedWAFAccessLogs(t *testing.T, ctx context.Context, loggedAt time.Time, remoteAddr string, host string, total int, notFound int) {
t.Helper()
records := make([]*model.OpenFlareAccessLog, 0, total)
for i := 0; i < total; i++ {
statusCode := http.StatusOK
if i < notFound {
statusCode = http.StatusNotFound
}
require.NoError(t, db.DB(ctx).Create(&model.OpenFlareAccessLog{
records = append(records, &model.OpenFlareAccessLog{
NodeID: "node-waf-auto",
LoggedAt: loggedAt.Add(-time.Duration(i%30) * time.Second),
RemoteAddr: remoteAddr,
Host: host,
Path: "/probe",
StatusCode: statusCode,
}).Error)
})
}
require.NoError(t, model.InsertOpenFlareAccessLogsBatch(ctx, records))
}