Files
seaweedfs/weed/filer/mysql/mysql_store.go
T
Chris Lu 9f6efc8b53 filer: a listing over a hard link no longer deadlocks a bounded SQL pool (#11118)
* filer: give the SQL stores' key-value reads their own connections

A listing holds the connection its rows are on for the whole iteration, and
FilerStoreWrapper calls maybeReadHardLink -> KvGet from inside that iteration,
so a hard-linked entry needs a second connection while the first is still busy.
Out of one bounded pool that is a deadlock: the listings fill the pool and then
wait for a connection none of them will release, and the wrapper's
context.WithoutCancel leaves the waiters without a deadline, so the filer stays
wedged rather than erroring.

The sqlite store shows it at its sharpest -- it allows a single connection, so
one listing over one hard-linked entry never returns. On postgres with
connection_max_open = 50, 60 concurrent listings over hard-linked entries made
no progress at all.

Key-value reads now run on their own pool, carved out of connection_max_open
rather than added to it, so the operator's cap still bounds what the store opens
against the database. An unbounded pool keeps a single pool: nothing can wait
there. sqlite's single connection becomes two, one per pool, and its writes get
a busy timeout so a write that meets the reader waits instead of failing.

Claude-Session: https://claude.ai/code/session_018DWwctzD4T2DmnczPRM47t

* sqlite: keep both pools on one database, whatever the dbFile spells

A dbFile that already carries URI options got a second "?" appended, which the
driver reads as part of the preceding option value, and a bare :memory: is
private to each connection, so the key-value pool would open its own empty
database and every key-value operation would fail on a missing filemeta.

Claude-Session: https://claude.ai/code/session_018DWwctzD4T2DmnczPRM47t

* sqlite: assert the busy timeout on the in-memory DSN too

Claude-Session: https://claude.ai/code/session_018DWwctzD4T2DmnczPRM47t
2026-09-02 23:45:45 -07:00

174 lines
5.5 KiB
Go

package mysql
import (
"crypto/tls"
"crypto/x509"
"database/sql"
"errors"
"fmt"
"os"
"strings"
"github.com/go-sql-driver/mysql"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/filer/abstract_sql"
"github.com/seaweedfs/seaweedfs/weed/util"
)
const (
CONNECTION_URL_PATTERN = "%s:%s@tcp(%s:%d)/%s?collation=utf8mb4_bin"
)
func init() {
filer.Stores = append(filer.Stores, &MysqlStore{})
}
type MysqlStore struct {
abstract_sql.AbstractSqlStore
}
func (store *MysqlStore) GetName() string {
return "mysql"
}
func (store *MysqlStore) Initialize(configuration util.Configuration, prefix string) (err error) {
// Fewer idle slots than concurrent operations means a fresh connection per
// operation, until the filer runs out of ephemeral ports. connection_max_open
// stays unset: a listing runs a second query from its own callback, so a
// bounded pool deadlocks once the concurrency reaches it.
configuration.SetDefault(prefix+"connection_max_idle", 50)
configuration.SetDefault(prefix+"connection_max_lifetime_seconds", 300)
// Default on so minimal configs avoid the duplicate-key roundtrip the
// inode-index KvPut would otherwise emit on every write.
configuration.SetDefault(prefix+"enableUpsert", true)
return store.initialize(
configuration.GetString(prefix+"dsn"),
configuration.GetString(prefix+"upsertQuery"),
configuration.GetBool(prefix+"enableUpsert"),
configuration.GetString(prefix+"username"),
configuration.GetString(prefix+"password"),
configuration.GetString(prefix+"hostname"),
configuration.GetInt(prefix+"port"),
configuration.GetString(prefix+"database"),
configuration.GetInt(prefix+"connection_max_idle"),
configuration.GetInt(prefix+"connection_max_open"),
configuration.GetInt(prefix+"connection_max_lifetime_seconds"),
configuration.GetBool(prefix+"interpolateParams"),
configuration.GetBool(prefix+"enable_tls"),
configuration.GetString(prefix+"ca_crt"),
configuration.GetString(prefix+"client_crt"),
configuration.GetString(prefix+"client_key"),
configuration.GetBool(prefix+"tls_insecure_skip_verify"),
configuration.GetString(prefix+"tls_server_name"),
)
}
func (store *MysqlStore) initialize(dsn string, upsertQuery string, enableUpsert bool, user, password, hostname string, port int, database string, maxIdle, maxOpen,
maxLifetimeSeconds int, interpolateParams bool, enableTls bool, caCrtDir string, clientCrtDir string, clientKeyDir string,
tlsInsecureSkipVerify bool, tlsServerName string) (err error) {
store.SupportBucketTable = false
if !enableUpsert {
upsertQuery = ""
} else if upsertQuery == "" {
upsertQuery = DefaultUpsertQuery
}
gen := &SqlGenMysql{
CreateTableSqlTemplate: "",
DropTableSqlTemplate: "DROP TABLE IF EXISTS `%s`",
UpsertQueryTemplate: upsertQuery,
}
store.SqlGenerator = gen
store.RetryableErrorCallback = func(err error) bool {
var mysqlError *mysql.MySQLError
if errors.As(err, &mysqlError) {
if mysqlError.Number == 1213 { // ER_LOCK_DEADLOCK
return true
}
if mysqlError.Number == 1205 { // ER_LOCK_WAIT_TIMEOUT
return true
}
}
return false
}
if dsn == "" {
dsn = fmt.Sprintf(CONNECTION_URL_PATTERN, user, password, hostname, port, database)
if interpolateParams {
dsn += "&interpolateParams=true"
}
}
cfg, err := mysql.ParseDSN(dsn)
if err != nil {
return fmt.Errorf("can not parse DSN error:%w", err)
}
if enableTls {
tlsConfig := &tls.Config{
MinVersion: tls.VersionTLS12,
InsecureSkipVerify: tlsInsecureSkipVerify,
ServerName: tlsServerName,
}
// When ca_crt is empty, leave RootCAs nil so Go falls back to the
// system trust store. This is the common case for managed databases
// (RDS, Aiven, ...) whose certs chain to a public CA already on the host.
if caCrtDir != "" {
rootCertPool := x509.NewCertPool()
pem, err := os.ReadFile(caCrtDir)
if err != nil {
return fmt.Errorf("read ca_crt %s: %w", caCrtDir, err)
}
if ok := rootCertPool.AppendCertsFromPEM(pem); !ok {
return fmt.Errorf("failed to append root certificate from %s", caCrtDir)
}
tlsConfig.RootCAs = rootCertPool
}
// Only attempt to load a client keypair when at least one of the paths is
// set. If either is set, both must load successfully — silently skipping
// a typo'd path used to mask broken mTLS setups as confusing handshake
// failures.
if clientCrtDir != "" || clientKeyDir != "" {
cert, err := tls.LoadX509KeyPair(clientCrtDir, clientKeyDir)
if err != nil {
return fmt.Errorf("load mysql client keypair (crt=%s key=%s): %w", clientCrtDir, clientKeyDir, err)
}
tlsConfig.Certificates = []tls.Certificate{cert}
}
// Set TLS directly on the parsed Config rather than registering a global
// "mysql-tls" entry — the global registry is process-wide and would be
// overwritten if a second MysqlStore is initialized with different TLS
// settings.
cfg.TLS = tlsConfig
}
connector, err := mysql.NewConnector(cfg)
if err != nil {
return fmt.Errorf("can not create mysql connector for %s error:%w", maskedDSN(cfg), err)
}
if err = store.UseConnectionPools(sql.OpenDB(connector), func() (*sql.DB, error) {
return sql.OpenDB(connector), nil
}, maxIdle, maxOpen, maxLifetimeSeconds); err != nil {
return err
}
if err = store.DB.Ping(); err != nil {
return fmt.Errorf("connect to %s error:%v", maskedDSN(cfg), err)
}
ConfigureListOrdering(store.DB, gen)
return nil
}
func maskedDSN(cfg *mysql.Config) string {
if cfg.Passwd == "" {
return cfg.FormatDSN()
}
return strings.ReplaceAll(cfg.FormatDSN(), cfg.Passwd, "<ADAPTED>")
}