BUG#35542 Add option to sync master and relay log to disk after every event

BUG#31665 sync_binlog should cause relay logs to be synchronized

NOTE: Backporting the patch to next-mr.
      
Add sync_relay_log option to server, this option works for relay log 
the same as option sync_binlog for binlog. This option also synchronize
master info to disk when set to non-zero value.
            
Original patches from Sinisa and Mark, with some modifications

sql/log.cc

@@ -49,8 +49,7 @@
 
 LOGGER logger;
 
-MYSQL_BIN_LOG mysql_bin_log;
+MYSQL_BIN_LOG mysql_bin_log(&sync_binlog_period);
-ulong sync_binlog_counter= 0;
 
 static bool test_if_number(const char *str,
 			   long *res, bool allow_wildcards);
@@ -2410,9 +2409,10 @@ const char *MYSQL_LOG::generate_name(const char *log_name,
 
 
 
-MYSQL_BIN_LOG::MYSQL_BIN_LOG()
+MYSQL_BIN_LOG::MYSQL_BIN_LOG(uint *sync_period)
   :bytes_written(0), prepared_xids(0), file_id(1), open_count(1),
    need_start_event(TRUE), m_table_map_version(0),
+   sync_period_ptr(sync_period),
    is_relay_log(0),
    description_event_for_exec(0), description_event_for_queue(0)
 {
@@ -3643,6 +3643,8 @@ bool MYSQL_BIN_LOG::append(Log_event* ev)
   }
   bytes_written+= ev->data_written;
   DBUG_PRINT("info",("max_size: %lu",max_size));
+  if (flush_and_sync(0))
+    goto err;
   if ((uint) my_b_append_tell(&log_file) > max_size)
     new_file_without_locking();
 
@@ -3673,6 +3675,8 @@ bool MYSQL_BIN_LOG::appendv(const char* buf, uint len,...)
     bytes_written += len;
   } while ((buf=va_arg(args,const char*)) && (len=va_arg(args,uint)));
   DBUG_PRINT("info",("max_size: %lu",max_size));
+  if (flush_and_sync(0))
+    goto err;
   if ((uint) my_b_append_tell(&log_file) > max_size)
     new_file_without_locking();
 
@@ -3682,17 +3686,21 @@ err:
   DBUG_RETURN(error);
 }
 
-
+bool MYSQL_BIN_LOG::flush_and_sync(bool *synced)
-bool MYSQL_BIN_LOG::flush_and_sync()
 {
   int err=0, fd=log_file.file;
+  if (synced)
+    *synced= 0;
   safe_mutex_assert_owner(&LOCK_log);
   if (flush_io_cache(&log_file))
     return 1;
-  if (++sync_binlog_counter >= sync_binlog_period && sync_binlog_period)
+  uint sync_period= get_sync_period();
+  if (sync_period && ++sync_counter >= sync_period)
   {
-    sync_binlog_counter= 0;
+    sync_counter= 0;
     err=my_sync(fd, MYF(MY_WME));
+    if (synced)
+      *synced= 1;
   }
   return err;
 }
@@ -3983,7 +3991,7 @@ MYSQL_BIN_LOG::flush_and_set_pending_rows_event(THD *thd,
 
     if (file == &log_file)
     {
-      error= flush_and_sync();
+      error= flush_and_sync(0);
       if (!error)
       {
         signal_update();
@@ -4169,7 +4177,8 @@ bool MYSQL_BIN_LOG::write(Log_event *event_info)
 
     if (file == &log_file) // we are writing to the real log (disk)
     {
-      if (flush_and_sync())
+      bool synced;
+      if (flush_and_sync(&synced))
 	goto err;
       signal_update();
       rotate_and_purge(RP_LOCK_LOG_IS_ALREADY_LOCKED);
@@ -4425,7 +4434,7 @@ int MYSQL_BIN_LOG::write_cache(IO_CACHE *cache, bool lock_log, bool sync_log)
   DBUG_ASSERT(carry == 0);
 
   if (sync_log)
-    flush_and_sync();
+    return flush_and_sync(0);
 
   return 0;                                     // All OK
 }
@@ -4472,7 +4481,8 @@ bool MYSQL_BIN_LOG::write_incident(THD *thd, bool lock)
   ev.write(&log_file);
   if (lock)
   {
-    if (!error && !(error= flush_and_sync()))
+    bool synced;
+    if (!error && !(error= flush_and_sync(&synced)))
     {
       signal_update();
       rotate_and_purge(RP_LOCK_LOG_IS_ALREADY_LOCKED);
@@ -4560,7 +4570,8 @@ bool MYSQL_BIN_LOG::write(THD *thd, IO_CACHE *cache, Log_event *commit_event,
       if (incident && write_incident(thd, FALSE))
         goto err;
 
-      if (flush_and_sync())
+      bool synced;
+      if (flush_and_sync(&synced))
         goto err;
       DBUG_EXECUTE_IF("half_binlogged_transaction", abort(););
       if (cache->error)				// Error on read

sql/log.h

@@ -269,6 +269,18 @@ class MYSQL_BIN_LOG: public TC_LOG, private MYSQL_LOG
 
   ulonglong m_table_map_version;
 
+  /* pointer to the sync period variable, for binlog this will be
+     sync_binlog_period, for relay log this will be
+     sync_relay_log_period
+  */
+  uint *sync_period_ptr;
+  uint sync_counter;
+
+  inline uint get_sync_period()
+  {
+    return *sync_period_ptr;
+  }
+
   int write_to_file(IO_CACHE *cache);
   /*
     This is used to start writing to a new log file. The difference from
@@ -296,7 +308,7 @@ public:
   Format_description_log_event *description_event_for_exec,
     *description_event_for_queue;
 
-  MYSQL_BIN_LOG();
+  MYSQL_BIN_LOG(uint *sync_period);
   /*
     note that there's no destructor ~MYSQL_BIN_LOG() !
     The reason is that we don't want it to be automatically called
@@ -378,7 +390,20 @@ public:
   bool is_active(const char* log_file_name);
   int update_log_index(LOG_INFO* linfo, bool need_update_threads);
   void rotate_and_purge(uint flags);
-  bool flush_and_sync();
+  /**
+     Flush binlog cache and synchronize to disk.
+
+     This function flushes events in binlog cache to binary log file,
+     it will do synchronizing according to the setting of system
+     variable 'sync_binlog'. If file is synchronized, @c synced will
+     be set to 1, otherwise 0.
+
+     @param[out] synced if not NULL, set to 1 if file is synchronized, otherwise 0
+
+     @retval 0 Success
+     @retval other Failure
+  */
+  bool flush_and_sync(bool *synced);
   int purge_logs(const char *to_log, bool included,
                  bool need_mutex, bool need_update_threads,
                  ulonglong *decrease_log_space);

sql/mysql_priv.h

@@ -1868,7 +1868,8 @@ extern ulong MYSQL_PLUGIN_IMPORT specialflag;
 #endif /* MYSQL_SERVER || INNODB_COMPATIBILITY_HOOKS */
 #ifdef MYSQL_SERVER
 extern ulong current_pid;
-extern ulong expire_logs_days, sync_binlog_period, sync_binlog_counter;
+extern ulong expire_logs_days;
+extern uint sync_binlog_period, sync_relaylog_period;
 extern ulong opt_tc_log_size, tc_log_max_pages_used, tc_log_page_size;
 extern ulong tc_log_page_waits;
 extern my_bool relay_log_purge, opt_innodb_safe_binlog, opt_innodb;

sql/mysqld.cc

@@ -552,7 +552,8 @@ ulong max_prepared_stmt_count;
 */
 ulong prepared_stmt_count=0;
 ulong thread_id=1L,current_pid;
-ulong slow_launch_threads = 0, sync_binlog_period;
+ulong slow_launch_threads = 0;
+uint sync_binlog_period= 0, sync_relaylog_period= 0;
 ulong expire_logs_days = 0;
 ulong rpl_recovery_rank=0;
 const char *log_output_str= "FILE";
@@ -5667,7 +5668,8 @@ enum options_mysqld
   OPT_SLAVE_EXEC_MODE,
   OPT_GENERAL_LOG_FILE,
   OPT_SLOW_QUERY_LOG_FILE,
-  OPT_IGNORE_BUILTIN_INNODB
+  OPT_IGNORE_BUILTIN_INNODB,
+  OPT_SYNC_RELAY_LOG
 };
 
 
@@ -6921,8 +6923,13 @@ The minimum value for this variable is 4096.",
   {"sync-binlog", OPT_SYNC_BINLOG,
    "Synchronously flush binary log to disk after every #th event. "
    "Use 0 (default) to disable synchronous flushing.",
-   (uchar**) &sync_binlog_period, (uchar**) &sync_binlog_period, 0, GET_ULONG,
+   (uchar**) &sync_binlog_period, (uchar**) &sync_binlog_period, 0, GET_UINT,
-   REQUIRED_ARG, 0, 0, ULONG_MAX, 0, 1, 0},
+   REQUIRED_ARG, 0, 0, (longlong) UINT_MAX, 0, 1, 0},
+  {"sync-relay-log", OPT_SYNC_RELAY_LOG,
+   "Synchronously flush relay log to disk after every #th event. "
+   "Use 0 (default) to disable synchronous flushing.",
+   (uchar**) &sync_relaylog_period, (uchar**) &sync_relaylog_period, 0, GET_UINT,
+   REQUIRED_ARG, 0, 0, (longlong) UINT_MAX, 0, 1, 0},
   {"sync-frm", OPT_SYNC_FRM, "Sync .frm to disk on create. Enabled by default.",
    (uchar**) &opt_sync_frm, (uchar**) &opt_sync_frm, 0, GET_BOOL, NO_ARG, 1, 0,
    0, 0, 0, 0},

sql/rpl_mi.cc

@@ -342,6 +342,7 @@ int flush_master_info(Master_info* mi, bool flush_relay_log_cache)
 {
   IO_CACHE* file = &mi->file;
   char lbuf[22];
+  int err= 0;
 
   DBUG_ENTER("flush_master_info");
   DBUG_PRINT("enter",("master_pos: %ld", (long) mi->master_log_pos));
@@ -358,9 +359,17 @@ int flush_master_info(Master_info* mi, bool flush_relay_log_cache)
     When we come to this place in code, relay log may or not be initialized;
     the caller is responsible for setting 'flush_relay_log_cache' accordingly.
   */
-  if (flush_relay_log_cache &&
+  if (flush_relay_log_cache)
-      flush_io_cache(mi->rli.relay_log.get_log_file()))
+  {
-    DBUG_RETURN(2);
+    IO_CACHE *log_file= mi->rli.relay_log.get_log_file();
+    if (flush_io_cache(log_file))
+      DBUG_RETURN(2);
+
+    /* Sync to disk if --sync-relay-log is set */
+    if (sync_relaylog_period &&
+        my_sync(log_file->file, MY_WME))
+      DBUG_RETURN(2);
+  }
 
   /*
     We flushed the relay log BEFORE the master.info file, because if we crash
@@ -388,7 +397,10 @@ int flush_master_info(Master_info* mi, bool flush_relay_log_cache)
               mi->password, mi->port, mi->connect_retry,
               (int)(mi->ssl), mi->ssl_ca, mi->ssl_capath, mi->ssl_cert,
               mi->ssl_cipher, mi->ssl_key, mi->ssl_verify_server_cert);
-  DBUG_RETURN(-flush_io_cache(file));
+  err= flush_io_cache(file);
+  if (sync_relaylog_period && !err)
+    err= my_sync(mi->fd, MYF(MY_WME));
+  DBUG_RETURN(-err);
 }
 
 

sql/rpl_rli.cc

@@ -32,7 +32,8 @@ int init_strvar_from_file(char *var, int max_size, IO_CACHE *f,
 Relay_log_info::Relay_log_info()
   :Slave_reporting_capability("SQL"),
    no_storage(FALSE), replicate_same_server_id(::replicate_same_server_id),
-   info_fd(-1), cur_log_fd(-1), save_temporary_tables(0),
+   info_fd(-1), cur_log_fd(-1), relay_log(&sync_relaylog_period),
+   save_temporary_tables(0),
 #if HAVE_purify
    is_fake(FALSE),
 #endif

sql/set_var.cc

@@ -1534,6 +1534,23 @@ static bool get_unsigned(THD *thd, set_var *var, ulonglong user_max,
 }
 
 
+bool sys_var_int_ptr::check(THD *thd, set_var *var)
+{
+  var->save_result.ulong_value= (ulong) var->value->val_int();
+  return 0;
+}
+
+bool sys_var_int_ptr::update(THD *thd, set_var *var)
+{
+  *value= (uint) var->save_result.ulong_value;
+  return 0;
+}
+
+void sys_var_int_ptr::set_default(THD *thd, enum_var_type type)
+{
+  *value= (uint) option_limits->def_value;
+}
+
 sys_var_long_ptr::
 sys_var_long_ptr(sys_var_chain *chain, const char *name_arg, ulong *value_ptr_arg,
                  sys_after_update_func after_update_arg)

sql/set_var.h

@@ -175,6 +175,27 @@ public:
   { return (uchar*) value; }
 };
 
+/**
+   Unsigned int system variable class
+ */
+class sys_var_int_ptr :public sys_var
+{
+public:
+  sys_var_int_ptr(sys_var_chain *chain, const char *name_arg, 
+                  uint *value_ptr_arg,
+                  sys_after_update_func after_update_arg= NULL)
+    :sys_var(name_arg, after_update_arg),
+     value(value_ptr_arg)
+  { chain_sys_var(chain); }
+  bool check(THD *thd, set_var *var);
+  bool update(THD *thd, set_var *var);
+  void set_default(THD *thd, enum_var_type type);
+  SHOW_TYPE show_type() { return SHOW_INT; }
+  uchar *value_ptr(THD *thd, enum_var_type type, LEX_STRING *base)
+  { return (uchar*) value; }
+private:
+  uint *value;
+};
 
 /*
   A global ulong variable that is protected by LOCK_global_system_variables

sql/sql_repl.cc

@@ -1753,15 +1753,6 @@ public:
   */
 };
 
-class sys_var_sync_binlog_period :public sys_var_long_ptr
-{
-public:
-  sys_var_sync_binlog_period(sys_var_chain *chain, const char *name_arg, 
-                             ulong *value_ptr)
-    :sys_var_long_ptr(chain, name_arg,value_ptr) {}
-  bool update(THD *thd, set_var *var);
-};
-
 static sys_var_chain vars = { NULL, NULL };
 
 static sys_var_const    sys_log_slave_updates(&vars, "log_slave_updates",
@@ -1793,7 +1784,8 @@ static sys_var_const    sys_slave_skip_errors(&vars, "slave_skip_errors",
                                               (uchar*) slave_skip_error_names);
 static sys_var_long_ptr	sys_slave_trans_retries(&vars, "slave_transaction_retries",
 						&slave_trans_retries);
-static sys_var_sync_binlog_period sys_sync_binlog_period(&vars, "sync_binlog", &sync_binlog_period);
+static sys_var_int_ptr sys_sync_binlog_period(&vars, "sync_binlog", &sync_binlog_period);
+static sys_var_int_ptr sys_sync_relaylog_period(&vars, "sync_relay_log", &sync_relaylog_period);
 static sys_var_slave_skip_counter sys_slave_skip_counter(&vars, "sql_slave_skip_counter");
 
 
@@ -1835,12 +1827,6 @@ bool sys_var_slave_skip_counter::update(THD *thd, set_var *var)
 }
 
 
-bool sys_var_sync_binlog_period::update(THD *thd, set_var *var)
-{
-  sync_binlog_period= (ulong) var->save_result.ulonglong_value;
-  return 0;
-}
-
 int init_replication_sys_vars()
 {
   if (mysql_add_sys_var_chain(vars.first, my_long_options))