[GITEA] Add slow SQL query warning

- Databases are one of the most important parts of Forgejo, every
interaction with Forgejo uses the database in one way or another.
Therefore, it is important to maintain the database and recognize when
Forgejo is not doing well with the database. Forgejo already has the
option to log *every* SQL query along with its execution time, but
monitoring becomes impractical for larger instances and takes up
unnecessary storage in the logs.
- Add a QoL enhancement that allows instance administrators to specify a
threshold value beyond which query execution time is logged as a warning
in the xorm logger. The default value is a conservative five seconds to
avoid this becoming a source of spam in the logs.
- The use case for this patch is that with an instance the size of Codeberg, monitoring SQL logs is not very fruitful and most of them are uninteresting. Recently, in the context of persistent deadlock issues (https://codeberg.org/forgejo/forgejo/issues/220), I have noticed that certain queries hold locks on tables like comment and issue for several seconds. This patch helps to identify which queries these are and when they happen.
- Added unit test.

(cherry picked from commit 24bbe7886f)
(cherry picked from commit 6e29145b3c)
(cherry picked from commit 63731e3071)
(cherry picked from commit 3ce1a09736)
(cherry picked from commit a64426907d)
(cherry picked from commit 4b19215691)
(cherry picked from commit e635674435)
(cherry picked from commit 9cf501f1af)
(cherry picked from commit 0d6b934eba)
(cherry picked from commit 4b6c273879)
(cherry picked from commit b50517139c)
(cherry picked from commit 6546dd1fc9)
(cherry picked from commit 3eda6890e6)

[GITEA] Add slow SQL query warning (squash) document the setting

(cherry picked from commit ce38599c51)
(cherry picked from commit 794aa67c68)
(cherry picked from commit 8227673deb)
(cherry picked from commit 8854d1d4dd)
(cherry picked from commit 9121a0e21f)
(cherry picked from commit 41bae2e425)

custom/conf/app.example.ini

@@ -409,6 +409,10 @@ USER = root
 ;;
 ;; Whether execute database models migrations automatically
 ;AUTO_MIGRATION = true
+;;
+;; Threshold value (in seconds) beyond which query execution time is logged as a warning in the xorm logger
+;;
+;SLOW_QUERY_TRESHOLD = 5s
 
 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

docs/content/administration/config-cheat-sheet.en-us.md

@@ -454,6 +454,7 @@ The following configuration set `Content-Type: application/vnd.android.package-a
 - `MAX_IDLE_CONNS` **2**: Max idle database connections on connection pool, default is 2 - this will be capped to `MAX_OPEN_CONNS`.
 - `CONN_MAX_LIFETIME` **0 or 3s**: Sets the maximum amount of time a DB connection may be reused - default is 0, meaning there is no limit (except on MySQL where it is 3s - see #6804 & #7071).
 - `AUTO_MIGRATION` **true**: Whether execute database models migrations automatically.
+- `SLOW_QUERY_TRESHOLD` **5s**: Threshold value in seconds beyond which query execution time is logged as a warning in the xorm logger.
 
 [^1]: It may be necessary to specify a hostport even when listening on a unix socket, as the port is part of the socket name. see [#24552](https://github.com/go-gitea/gitea/issues/24552#issuecomment-1681649367) for additional details.
 

models/db/engine.go

@@ -11,10 +11,13 @@ import (
 	"io"
 	"reflect"
 	"strings"
+	"time"
 
+	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
 
 	"xorm.io/xorm"
+	"xorm.io/xorm/contexts"
 	"xorm.io/xorm/names"
 	"xorm.io/xorm/schemas"
 
@@ -147,6 +150,13 @@ func InitEngine(ctx context.Context) error {
 	xormEngine.SetConnMaxLifetime(setting.Database.ConnMaxLifetime)
 	xormEngine.SetDefaultContext(ctx)
 
+	if setting.Database.SlowQueryTreshold > 0 {
+		xormEngine.AddHook(&SlowQueryHook{
+			Treshold: setting.Database.SlowQueryTreshold,
+			Logger:   log.GetLogger("xorm"),
+		})
+	}
+
 	SetDefaultEngine(ctx, xormEngine)
 	return nil
 }
@@ -300,3 +310,21 @@ func SetLogSQL(ctx context.Context, on bool) {
 		sess.Engine().ShowSQL(on)
 	}
 }
+
+type SlowQueryHook struct {
+	Treshold time.Duration
+	Logger   log.Logger
+}
+
+var _ contexts.Hook = &SlowQueryHook{}
+
+func (SlowQueryHook) BeforeProcess(c *contexts.ContextHook) (context.Context, error) {
+	return c.Ctx, nil
+}
+
+func (h *SlowQueryHook) AfterProcess(c *contexts.ContextHook) error {
+	if c.ExecuteTime >= h.Treshold {
+		h.Logger.Log(8, log.WARN, "[Slow SQL Query] %s %v - %v", c.SQL, c.Args, c.ExecuteTime)
+	}
+	return nil
+}

models/db/engine_test.go

@@ -6,15 +6,19 @@ package db_test
 import (
 	"path/filepath"
 	"testing"
+	"time"
 
 	"code.gitea.io/gitea/models/db"
 	issues_model "code.gitea.io/gitea/models/issues"
 	"code.gitea.io/gitea/models/unittest"
+	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
+	"code.gitea.io/gitea/modules/test"
 
 	_ "code.gitea.io/gitea/cmd" // for TestPrimaryKeys
 
 	"github.com/stretchr/testify/assert"
+	"xorm.io/xorm"
 )
 
 func TestDumpDatabase(t *testing.T) {
@@ -85,3 +89,37 @@ func TestPrimaryKeys(t *testing.T) {
 		}
 	}
 }
+
+func TestSlowQuery(t *testing.T) {
+	lc, cleanup := test.NewLogChecker("slow-query")
+	lc.StopMark("[Slow SQL Query]")
+	defer cleanup()
+
+	e := db.GetEngine(db.DefaultContext)
+	engine, ok := e.(*xorm.Engine)
+	assert.True(t, ok)
+
+	// It's not possible to clean this up with XORM, but it's luckily not harmful
+	// to leave around.
+	engine.AddHook(&db.SlowQueryHook{
+		Treshold: time.Second * 10,
+		Logger:   log.GetLogger("slow-query"),
+	})
+
+	// NOOP query.
+	e.Exec("SELECT 1 WHERE false;")
+
+	_, stopped := lc.Check(100 * time.Millisecond)
+	assert.False(t, stopped)
+
+	engine.AddHook(&db.SlowQueryHook{
+		Treshold: 0, // Every query should be logged.
+		Logger:   log.GetLogger("slow-query"),
+	})
+
+	// NOOP query.
+	e.Exec("SELECT 1 WHERE false;")
+
+	_, stopped = lc.Check(100 * time.Millisecond)
+	assert.True(t, stopped)
+}

modules/setting/database.go

@@ -45,6 +45,7 @@ var (
 		ConnMaxLifetime   time.Duration
 		IterateBufferSize int
 		AutoMigration     bool
+		SlowQueryTreshold time.Duration
 	}{
 		Timeout:           500,
 		IterateBufferSize: 50,
@@ -87,6 +88,7 @@ func loadDBSetting(rootCfg ConfigProvider) {
 	Database.DBConnectRetries = sec.Key("DB_RETRIES").MustInt(10)
 	Database.DBConnectBackoff = sec.Key("DB_RETRY_BACKOFF").MustDuration(3 * time.Second)
 	Database.AutoMigration = sec.Key("AUTO_MIGRATION").MustBool(true)
+	Database.SlowQueryTreshold = sec.Key("SLOW_QUERY_TRESHOLD").MustDuration(5 * time.Second)
 }
 
 // DBConnStr returns database connection string