diff --git a/source/dnode/vnode/src/meta/metaSnapshot.c b/source/dnode/vnode/src/meta/metaSnapshot.c index bb20a1a7ff..7f69c7a638 100644 --- a/source/dnode/vnode/src/meta/metaSnapshot.c +++ b/source/dnode/vnode/src/meta/metaSnapshot.c @@ -194,4 +194,4 @@ int32_t metaSnapWrite(SMetaSnapWriter* pWriter, uint8_t* pData, uint32_t nData) _err: metaError("vgId:%d vnode snapshot meta write failed since %s", TD_VID(pMeta->pVnode), tstrerror(code)); return code; -} \ No newline at end of file +} diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas_querys.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas_querys.py new file mode 100644 index 0000000000..02d944b08f --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas_querys.py @@ -0,0 +1,202 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 1 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 1000 + self.query_times = 500 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + ts = self.ts + 1000*row_num + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + + + def check_insert_status(self, newTdSql ,dbname, tb_nums , row_nums): + # newTdSql=tdCom.newTdSql() + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + self.check_insert_status( newTdSql , db_name, tb_nums , row_nums) + + def run(self): + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + reading = threading.Thread(target = self.loop_query_constantly, args=(self.query_times,self.db_name , self.tb_nums , self.row_nums)) + reading.start() + + writing.join() + reading.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_force_stop_all_dnodes.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_force_stop_all_dnodes.py new file mode 100644 index 0000000000..3d01015af6 --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_force_stop_all_dnodes.py @@ -0,0 +1,428 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import datetime +import inspect +import time +import socket +import subprocess +import threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.ts_step =1000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 100 + self.stop_dnode_id = None + self.loop_restart_times = 5 + self.current_thread = None + self.max_restart_time = 10 + self.try_check_times = 10 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_database(self, dbname, replica_num ,vgroup_nums ): + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + tdSql.execute(drop_db_sql) + tdSql.execute(create_db_sql) + tdSql.execute("use {}".format(dbname)) + + def create_stable_insert_datas(self,dbname ,stablename , tb_nums , row_nums): + tdSql.execute("use {}".format(dbname)) + tdSql.execute( + '''create table {} + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + '''.format(stablename) + ) + + for i in range(tb_nums): + sub_tbname = "sub_{}_{}".format(stablename,i) + tdSql.execute("create table {} using {} tags({})".format(sub_tbname, stablename ,i)) + # insert datas about new database + + for row_num in range(row_nums): + ts = self.ts + self.ts_step*row_num + tdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== stable {} insert rows execute end =====".format(stablename)) + + def append_rows_of_exists_tables(self,dbname ,stablename , tbname , append_nums ): + + tdSql.execute("use {}".format(dbname)) + + for row_num in range(append_nums): + tdSql.execute(f"insert into {tbname} values (now, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + # print(f"insert into {tbname} values (now, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + tdLog.notice(" ==== append new rows of table {} belongs to stable {} execute end =====".format(tbname,stablename)) + os.system("taos -s 'select count(*) from {}.{}';".format(dbname,stablename)) + + def check_insert_rows(self, dbname, stablename , tb_nums , row_nums, append_rows): + + tdSql.execute("use {}".format(dbname)) + + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + + status_OK = self.mycheckData("select count(*) from {}.{}".format(dbname,stablename) ,0 , 0 , tb_nums*row_nums+append_rows) + + count = 0 + while not status_OK : + if count > self.try_check_times: + os.system("taos -s ' show {}.vgroups; '".format(dbname)) + tdLog.exit(" ==== check insert rows failed after {} try check {} times of database {}".format(count , self.try_check_times ,dbname)) + break + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckData("select count(*) from {}.{}".format(dbname,stablename) ,0 , 0 , tb_nums*row_nums+append_rows) + tdLog.debug(" ==== check insert rows first failed , this is {}_th retry check rows of database {}".format(count , dbname)) + count += 1 + + + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) + count = 0 + while not status_OK : + if count > self.try_check_times: + os.system("taos -s ' show {}.vgroups;'".format(dbname)) + tdLog.exit(" ==== check insert rows failed after {} try check {} times of database {}".format(count , self.try_check_times ,dbname)) + break + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) + tdLog.debug(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) + count += 1 + + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) + for vgroup_info in vgroup_infos: + leader_infos = vgroup_info[3:-4] + # print(vgroup_info) + for ind ,role in enumerate(leader_infos): + if role == dnode_role: + # print(ind,leader_infos) + self.stop_dnode_id = leader_infos[ind-1] + break + + return self.stop_dnode_id + + def wait_stop_dnode_OK(self ,newTdSql ): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="offline": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) + + def wait_start_dnode_OK(self , newTdSql): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="ready": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has restart , id is {} ====".format(self.stop_dnode_id)) + + def _parse_datetime(self,timestr): + try: + return datetime.datetime.strptime(timestr, '%Y-%m-%d %H:%M:%S.%f') + except ValueError: + pass + try: + return datetime.datetime.strptime(timestr, '%Y-%m-%d %H:%M:%S') + except ValueError: + pass + + def mycheckRowCol(self, sql, row, col): + caller = inspect.getframeinfo(inspect.stack()[2][0]) + if row < 0: + args = (caller.filename, caller.lineno, sql, row) + tdLog.exit("%s(%d) failed: sql:%s, row:%d is smaller than zero" % args) + if col < 0: + args = (caller.filename, caller.lineno, sql, row) + tdLog.exit("%s(%d) failed: sql:%s, col:%d is smaller than zero" % args) + if row > tdSql.queryRows: + args = (caller.filename, caller.lineno, sql, row, tdSql.queryRows) + tdLog.exit("%s(%d) failed: sql:%s, row:%d is larger than queryRows:%d" % args) + if col > tdSql.queryCols: + args = (caller.filename, caller.lineno, sql, col, tdSql.queryCols) + tdLog.exit("%s(%d) failed: sql:%s, col:%d is larger than queryCols:%d" % args) + + def mycheckData(self, sql ,row, col, data): + check_status = True + self.mycheckRowCol(sql ,row, col) + if tdSql.queryResult[row][col] != data: + if tdSql.cursor.istype(col, "TIMESTAMP"): + # suppose user want to check nanosecond timestamp if a longer data passed + if (len(data) >= 28): + if pd.to_datetime(tdSql.queryResult[row][col]) == pd.to_datetime(data): + tdLog.info("sql:%s, row:%d col:%d data:%d == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + else: + if tdSql.queryResult[row][col] == self._parse_datetime(data): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + + if str(tdSql.queryResult[row][col]) == str(data): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + elif isinstance(data, float) and abs(tdSql.queryResult[row][col] - data) <= 0.000001: + tdLog.info("sql:%s, row:%d col:%d data:%f == expect:%f" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + else: + caller = inspect.getframeinfo(inspect.stack()[1][0]) + args = (caller.filename, caller.lineno, sql, row, col, tdSql.queryResult[row][col], data) + tdLog.info("%s(%d) failed: sql:%s row:%d col:%d data:%s != expect:%s" % args) + + check_status = False + + if data is None: + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + elif isinstance(data, str): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + # elif isinstance(data, datetime.date): + # tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + # (sql, row, col, tdSql.queryResult[row][col], data)) + elif isinstance(data, float): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + else: + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%d" % + (sql, row, col, tdSql.queryResult[row][col], data)) + + return check_status + + def mycheckRows(self, sql, expectRows): + check_status = True + if len(tdSql.queryResult) == expectRows: + tdLog.info("sql:%s, queryRows:%d == expect:%d" % (sql, len(tdSql.queryResult), expectRows)) + return True + else: + caller = inspect.getframeinfo(inspect.stack()[1][0]) + args = (caller.filename, caller.lineno, sql, len(tdSql.queryResult), expectRows) + tdLog.info("%s(%d) failed: sql:%s, queryRows:%d != expect:%d" % args) + check_status = False + return check_status + + def force_stop_dnode(self, dnode_id ): + + port = None + for k ,v in self.dnode_list.items(): + if v[0] == dnode_id: + port = k.split(":")[-1] + else: + continue + + if port: + tdLog.notice(" ==== dnode {} will be force stop by kill -9 ====".format(dnode_id)) + psCmd = '''netstat -anp|grep -w LISTEN|grep -w %s |grep -o "LISTEN.*"|awk '{print $2}'|cut -d/ -f1|head -n1''' %(port) + processID = subprocess.check_output( + psCmd, shell=True).decode("utf-8") + ps_kill_taosd = ''' kill -9 {} '''.format(processID) + # print(ps_kill_taosd) + os.system(ps_kill_taosd) + else : + tdLog.exit(" ==== port of dnode {} not found ====".format(dnode_id)) + + def stop_All(self): + + tdDnodes = cluster.dnodes + # newTdSql=tdCom.newTdSql() + # ==== stop all dnode ===== + for k ,v in self.dnode_list.items(): + dnode_id = v[0] + # tdDnodes[dnode_id-1].stoptaosd() + self.force_stop_dnode(dnode_id) + # self.wait_stop_dnode_OK(newTdSql) + + def start_All(self): + tdDnodes = cluster.dnodes + # newTdSql=tdCom.newTdSql() + for k ,v in self.dnode_list.items(): + dnode_id = v[0] + start = time.time() + tdDnodes[dnode_id-1].starttaosd() + # self.wait_start_dnode_OK(newTdSql) + end = time.time() + time_cost = int(end -start) + if time_cost > self.max_restart_time: + tdLog.exit(" ==== restart dnode {} cost too much time , please check ====".format(self.stop_dnode_id)) + + + + def stop_All_dnodes_check_datas(self): + + # stop follower and insert datas , update tables and create new stables + self.create_database(dbname = self.db_name ,replica_num= self.replica , vgroup_nums= 1) + self.create_stable_insert_datas(dbname = self.db_name , stablename = "stb1" , tb_nums= self.tb_nums ,row_nums= self.row_nums ) + self.check_insert_rows(self.db_name ,'stb1' ,tb_nums = self.tb_nums , row_nums= self.row_nums ,append_rows=0) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + for i in range(self.loop_restart_times): + # begin to stop All taosd + self.stop_All() + # begin to start All taosd + self.start_All() + + tdLog.debug(" ==== cluster has restart , this is {}_th restart cluster ==== ".format(i)) + + self.check_insert_rows(self.db_name ,'stb1' ,tb_nums = self.tb_nums , row_nums= self.row_nums ,append_rows=0) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + self.append_rows_of_exists_tables(self.db_name ,"stb1" , "sub_stb1_0" , 100 ) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + + def run(self): + + # basic insert and check of cluster + self.check_setup_cluster_status() + self.stop_All_dnodes_check_datas() + + + + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys.py new file mode 100644 index 0000000000..3649617c21 --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys.py @@ -0,0 +1,204 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 1000 + self.query_times = 1000 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + if row_num %100 ==0: + tdLog.info(" === writing is going now === ") + ts = self.ts + 1000*row_num + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + + + def check_insert_status(self, newTdSql ,dbname, tb_nums , row_nums): + + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + def run(self): + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + reading = threading.Thread(target = self.loop_query_constantly, args=(self.query_times,self.db_name , self.tb_nums , self.row_nums)) + reading.start() + + writing.join() + reading.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_all_vnode.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_all_vnode.py new file mode 100644 index 0000000000..db05eca9ce --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_all_vnode.py @@ -0,0 +1,355 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 500 + self.max_restart_time = 20 + self.restart_server_times = 10 + self.dnode_index = 0 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + if row_num % (int(row_nums*0.1)) == 0 : + tdLog.notice( " === database {} writing records {} rows".format(dbname , row_num ) ) + ts = self.ts + 1000*row_num + + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + + def _get_stop_dnode_id(self): + + + dnode_lists = list(set(self.dnode_list.keys()) -set(self.mnode_list.keys())) + # print(dnode_lists) + self.stop_dnode_id = self.dnode_list[dnode_lists[self.dnode_index % 3]][0] + + self.dnode_index += 1 + + return self.stop_dnode_id + + def wait_stop_dnode_OK(self , newTdSql): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + + while status !="offline": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) + + def wait_start_dnode_OK(self , newTdSql ): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="ready": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has restart , id is {} ====".format(self.stop_dnode_id)) + + def get_leader_infos(self , newTdSql , dbname): + + # newTdSql=tdCom.newTdSql() + newTdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = newTdSql.queryResult + + leader_infos = set() + for vgroup_info in vgroup_infos: + leader_infos.add(vgroup_info[3:-4]) + + return leader_infos + + def check_revote_leader_success(self, dbname, before_leader_infos , after_leader_infos): + + check_status = False + vote_act = set(set(after_leader_infos)-set(before_leader_infos)) + if not vote_act: + print("=======before_revote_leader_infos ======\n" , before_leader_infos) + print("=======after_revote_leader_infos ======\n" , after_leader_infos) + tdLog.exit(" ===maybe revote not occured , there is no dnode offline ====") + else: + for vgroup_info in vote_act: + for ind , role in enumerate(vgroup_info): + if role==self.stop_dnode_id: + + if vgroup_info[ind+1] =="offline" and "leader" in vgroup_info: + tdLog.notice(" === revote leader ok , leader is {} now ====".format(vgroup_info[list(vgroup_info).index("leader")-1])) + check_status = True + elif vgroup_info[ind+1] !="offline": + tdLog.notice(" === dnode {} should be offline ".format(self.stop_dnode_id)) + else: + continue + break + return check_status + + + def check_insert_status(self, newTdSql , dbname, tb_nums , row_nums): + + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + + def loop_restart_follower_constantly(self, times , db_name): + + tdDnodes = cluster.dnodes + newTdSql=tdCom.newTdSql() + + for loop_time in range(times): + + self.stop_dnode_id = self._get_stop_dnode_id() + + # print(self.stop_dnode_id) + # begin stop dnode + + # before_leader_infos = self.get_leader_infos( newTdSql ,db_name) + tdDnodes[self.stop_dnode_id-1].stoptaosd() + self.wait_stop_dnode_OK(newTdSql) + + # start = time.time() + # # get leader info after stop + # after_leader_infos = self.get_leader_infos(newTdSql , db_name) + + # revote_status = self.check_revote_leader_success(db_name ,before_leader_infos , after_leader_infos) + + # # append rows of stablename when dnode stop make sure revote leaders + + # while not revote_status: + # after_leader_infos = self.get_leader_infos(newTdSql , db_name) + # revote_status = self.check_revote_leader_success(db_name ,before_leader_infos , after_leader_infos) + + # end = time.time() + # time_cost = end - start + # tdLog.notice(" ==== revote leader of database {} cost time {} ====".format(db_name , time_cost)) + + tdLog.notice(" === this is {}_th restart taosd === ".format(loop_time)) + + # begin start dnode + start = time.time() + tdDnodes[self.stop_dnode_id-1].starttaosd() + self.wait_start_dnode_OK(newTdSql) + time.sleep(5) + end = time.time() + time_cost = int(end -start) + if time_cost > self.max_restart_time: + tdLog.exit(" ==== restart dnode {} cost too much time , please check ====".format(self.stop_dnode_id)) + + + + def run(self): + + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + restart_servers = threading.Thread(target = self.loop_restart_follower_constantly, args = (self.restart_server_times ,self.db_name)) + restart_servers.start() + + # reading = threading.Thread(target = self.loop_query_constantly, args=(1000,self.db_name , self.tb_nums , self.row_nums)) + # reading.start() + + writing.join() + # reading.join() + restart_servers.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_follower.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_follower.py new file mode 100644 index 0000000000..fdd5ec7d46 --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_follower.py @@ -0,0 +1,316 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 10 + self.tb_nums = 10 + self.row_nums = 100 + self.max_restart_time = 20 + self.restart_server_times = 5 + self.query_times = 100 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + if row_num % (int(row_nums*0.1)) == 0 : + tdLog.notice( " === database {} writing records {} rows".format(dbname , row_num ) ) + ts = self.ts + 1000*row_num + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) + for vgroup_info in vgroup_infos: + leader_infos = vgroup_info[3:-4] + # print(vgroup_info) + for ind ,role in enumerate(leader_infos): + if role == dnode_role: + # print(ind,leader_infos) + self.stop_dnode_id = leader_infos[ind-1] + break + + return self.stop_dnode_id + + def wait_stop_dnode_OK(self , newTdSql): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + + while status !="offline": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) + + def wait_start_dnode_OK(self , newTdSql ): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="ready": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has restart , id is {} ====".format(self.stop_dnode_id)) + + def check_insert_status(self, newTdSql , dbname, tb_nums , row_nums): + + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + + def loop_restart_follower_constantly(self, times , db_name): + + tdDnodes = cluster.dnodes + newTdSql=tdCom.newTdSql() + + for loop_time in range(times): + + self.stop_dnode_id = self._get_stop_dnode_id(db_name , "follower") + + print(self.stop_dnode_id) + # begin stop dnode + start = time.time() + tdDnodes[self.stop_dnode_id-1].stoptaosd() + self.wait_stop_dnode_OK(newTdSql) + + tdLog.notice(" === this is {}_th restart taosd === ".format(loop_time)) + + # begin start dnode + tdDnodes[self.stop_dnode_id-1].starttaosd() + self.wait_start_dnode_OK(newTdSql) + end = time.time() + time_cost = int(end -start) + if time_cost > self.max_restart_time: + tdLog.exit(" ==== restart dnode {} cost too much time , please check ====".format(self.stop_dnode_id)) + + + + def run(self): + + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + restart_servers = threading.Thread(target = self.loop_restart_follower_constantly, args = (self.restart_server_times ,self.db_name)) + restart_servers.start() + + reading = threading.Thread(target = self.loop_query_constantly, args=(self.query_times,self.db_name , self.tb_nums , self.row_nums)) + reading.start() + + writing.join() + reading.join() + restart_servers.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_leader.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_leader.py new file mode 100644 index 0000000000..cbb007d961 --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_leader.py @@ -0,0 +1,369 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 10 + self.tb_nums = 10 + self.row_nums = 100 + self.max_restart_time = 20 + self.restart_server_times = 10 + self.query_times = 100 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + if row_num % (int(row_nums*0.1)) == 0 : + tdLog.notice( " === database {} writing records {} rows".format(dbname , row_num ) ) + ts = self.ts + 1000*row_num + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) + for vgroup_info in vgroup_infos: + leader_infos = vgroup_info[3:-4] + # print(vgroup_info) + for ind ,role in enumerate(leader_infos): + if role == dnode_role: + # print(ind,leader_infos) + self.stop_dnode_id = leader_infos[ind-1] + break + + return self.stop_dnode_id + + def wait_stop_dnode_OK(self , newTdSql): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + + while status !="offline": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) + + def wait_start_dnode_OK(self , newTdSql ): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="ready": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has restart , id is {} ====".format(self.stop_dnode_id)) + + def get_leader_infos(self , newTdSql , dbname): + + # newTdSql=tdCom.newTdSql() + newTdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = newTdSql.queryResult + + leader_infos = set() + for vgroup_info in vgroup_infos: + leader_infos.add(vgroup_info[3:-4]) + + return leader_infos + + def check_revote_leader_success(self, dbname, before_leader_infos , after_leader_infos): + + check_status = False + vote_act = set(set(after_leader_infos)-set(before_leader_infos)) + if not vote_act: + print("=======before_revote_leader_infos ======\n" , before_leader_infos) + print("=======after_revote_leader_infos ======\n" , after_leader_infos) + tdLog.exit(" ===maybe revote not occured , there is no dnode offline ====") + else: + for vgroup_info in vote_act: + for ind , role in enumerate(vgroup_info): + if role==self.stop_dnode_id: + + if vgroup_info[ind+1] =="offline" and "leader" in vgroup_info: + tdLog.notice(" === revote leader ok , leader is {} now ====".format(vgroup_info[list(vgroup_info).index("leader")-1])) + check_status = True + elif vgroup_info[ind+1] !="offline": + tdLog.notice(" === dnode {} should be offline ".format(self.stop_dnode_id)) + else: + continue + break + return check_status + + + def check_insert_status(self, newTdSql , dbname, tb_nums , row_nums): + + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + + def loop_restart_follower_constantly(self, times , db_name): + + tdDnodes = cluster.dnodes + newTdSql=tdCom.newTdSql() + + for loop_time in range(times): + + self.stop_dnode_id = self._get_stop_dnode_id(db_name , "leader") + + # print(self.stop_dnode_id) + # begin stop dnode + start = time.time() + + before_leader_infos = self.get_leader_infos( newTdSql ,db_name) + tdDnodes[self.stop_dnode_id-1].stoptaosd() + self.wait_stop_dnode_OK(newTdSql) + + start = time.time() + # get leader info after stop + after_leader_infos = self.get_leader_infos(newTdSql , db_name) + + revote_status = self.check_revote_leader_success(db_name ,before_leader_infos , after_leader_infos) + + # append rows of stablename when dnode stop make sure revote leaders + + while not revote_status: + after_leader_infos = self.get_leader_infos(newTdSql , db_name) + revote_status = self.check_revote_leader_success(db_name ,before_leader_infos , after_leader_infos) + + end = time.time() + time_cost = end - start + tdLog.notice(" ==== revote leader of database {} cost time {} ====".format(db_name , time_cost)) + + tdLog.notice(" === this is {}_th restart taosd === ".format(loop_time)) + + # begin start dnode + tdDnodes[self.stop_dnode_id-1].starttaosd() + self.wait_start_dnode_OK(newTdSql) + end = time.time() + time_cost = int(end -start) + if time_cost > self.max_restart_time: + tdLog.exit(" ==== restart dnode {} cost too much time , please check ====".format(self.stop_dnode_id)) + + + + def run(self): + + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + restart_servers = threading.Thread(target = self.loop_restart_follower_constantly, args = (self.restart_server_times ,self.db_name)) + restart_servers.start() + + reading = threading.Thread(target = self.loop_query_constantly, args=(self.query_times,self.db_name , self.tb_nums , self.row_nums)) + reading.start() + + writing.join() + reading.join() + restart_servers.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_all_dnodes.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_all_dnodes.py new file mode 100644 index 0000000000..63c4942c9e --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_all_dnodes.py @@ -0,0 +1,407 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import datetime +import inspect +import time +import socket +import subprocess +import threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.ts_step =1000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 100 + self.stop_dnode_id = None + self.loop_restart_times = 5 + self.current_thread = None + self.max_restart_time = 10 + self.try_check_times = 10 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_database(self, dbname, replica_num ,vgroup_nums ): + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + tdSql.execute(drop_db_sql) + tdSql.execute(create_db_sql) + tdSql.execute("use {}".format(dbname)) + + def create_stable_insert_datas(self,dbname ,stablename , tb_nums , row_nums): + tdSql.execute("use {}".format(dbname)) + tdSql.execute( + '''create table {} + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + '''.format(stablename) + ) + + for i in range(tb_nums): + sub_tbname = "sub_{}_{}".format(stablename,i) + tdSql.execute("create table {} using {} tags({})".format(sub_tbname, stablename ,i)) + # insert datas about new database + + for row_num in range(row_nums): + ts = self.ts + self.ts_step*row_num + tdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== stable {} insert rows execute end =====".format(stablename)) + + def append_rows_of_exists_tables(self,dbname ,stablename , tbname , append_nums ): + + tdSql.execute("use {}".format(dbname)) + + for row_num in range(append_nums): + tdSql.execute(f"insert into {tbname} values (now, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + # print(f"insert into {tbname} values (now, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + tdLog.notice(" ==== append new rows of table {} belongs to stable {} execute end =====".format(tbname,stablename)) + os.system("taos -s 'select count(*) from {}.{}';".format(dbname,stablename)) + + def check_insert_rows(self, dbname, stablename , tb_nums , row_nums, append_rows): + + tdSql.execute("use {}".format(dbname)) + + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + + status_OK = self.mycheckData("select count(*) from {}.{}".format(dbname,stablename) ,0 , 0 , tb_nums*row_nums+append_rows) + + count = 0 + while not status_OK : + if count > self.try_check_times: + os.system("taos -s ' show {}.vgroups; '".format(dbname)) + tdLog.exit(" ==== check insert rows failed after {} try check {} times of database {}".format(count , self.try_check_times ,dbname)) + break + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select count(*) from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckData("select count(*) from {}.{}".format(dbname,stablename) ,0 , 0 , tb_nums*row_nums+append_rows) + tdLog.debug(" ==== check insert rows first failed , this is {}_th retry check rows of database {}".format(count , dbname)) + count += 1 + + + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) + count = 0 + while not status_OK : + if count > self.try_check_times: + os.system("taos -s ' show {}.vgroups;'".format(dbname)) + tdLog.exit(" ==== check insert rows failed after {} try check {} times of database {}".format(count , self.try_check_times ,dbname)) + break + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + while not tdSql.queryResult: + time.sleep(0.1) + tdSql.query("select distinct tbname from {}.{}".format(dbname,stablename)) + status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) + tdLog.debug(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) + count += 1 + + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) + for vgroup_info in vgroup_infos: + leader_infos = vgroup_info[3:-4] + # print(vgroup_info) + for ind ,role in enumerate(leader_infos): + if role == dnode_role: + # print(ind,leader_infos) + self.stop_dnode_id = leader_infos[ind-1] + break + + return self.stop_dnode_id + + def wait_stop_dnode_OK(self ,newTdSql ): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="offline": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) + + def wait_start_dnode_OK(self , newTdSql): + + def _get_status(): + # newTdSql=tdCom.newTdSql() + status = "" + newTdSql.query("show dnodes") + dnode_infos = newTdSql.queryResult + for dnode_info in dnode_infos: + id = dnode_info[0] + dnode_status = dnode_info[4] + if id == self.stop_dnode_id: + status = dnode_status + break + return status + + status = _get_status() + while status !="ready": + time.sleep(0.1) + status = _get_status() + # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) + tdLog.notice("==== stop_dnode has restart , id is {} ====".format(self.stop_dnode_id)) + + def _parse_datetime(self,timestr): + try: + return datetime.datetime.strptime(timestr, '%Y-%m-%d %H:%M:%S.%f') + except ValueError: + pass + try: + return datetime.datetime.strptime(timestr, '%Y-%m-%d %H:%M:%S') + except ValueError: + pass + + def mycheckRowCol(self, sql, row, col): + caller = inspect.getframeinfo(inspect.stack()[2][0]) + if row < 0: + args = (caller.filename, caller.lineno, sql, row) + tdLog.exit("%s(%d) failed: sql:%s, row:%d is smaller than zero" % args) + if col < 0: + args = (caller.filename, caller.lineno, sql, row) + tdLog.exit("%s(%d) failed: sql:%s, col:%d is smaller than zero" % args) + if row > tdSql.queryRows: + args = (caller.filename, caller.lineno, sql, row, tdSql.queryRows) + tdLog.exit("%s(%d) failed: sql:%s, row:%d is larger than queryRows:%d" % args) + if col > tdSql.queryCols: + args = (caller.filename, caller.lineno, sql, col, tdSql.queryCols) + tdLog.exit("%s(%d) failed: sql:%s, col:%d is larger than queryCols:%d" % args) + + def mycheckData(self, sql ,row, col, data): + check_status = True + self.mycheckRowCol(sql ,row, col) + if tdSql.queryResult[row][col] != data: + if tdSql.cursor.istype(col, "TIMESTAMP"): + # suppose user want to check nanosecond timestamp if a longer data passed + if (len(data) >= 28): + if pd.to_datetime(tdSql.queryResult[row][col]) == pd.to_datetime(data): + tdLog.info("sql:%s, row:%d col:%d data:%d == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + else: + if tdSql.queryResult[row][col] == self._parse_datetime(data): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + + if str(tdSql.queryResult[row][col]) == str(data): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + elif isinstance(data, float) and abs(tdSql.queryResult[row][col] - data) <= 0.000001: + tdLog.info("sql:%s, row:%d col:%d data:%f == expect:%f" % + (sql, row, col, tdSql.queryResult[row][col], data)) + return + else: + caller = inspect.getframeinfo(inspect.stack()[1][0]) + args = (caller.filename, caller.lineno, sql, row, col, tdSql.queryResult[row][col], data) + tdLog.info("%s(%d) failed: sql:%s row:%d col:%d data:%s != expect:%s" % args) + + check_status = False + + if data is None: + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + elif isinstance(data, str): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + # elif isinstance(data, datetime.date): + # tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + # (sql, row, col, tdSql.queryResult[row][col], data)) + elif isinstance(data, float): + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%s" % + (sql, row, col, tdSql.queryResult[row][col], data)) + else: + tdLog.info("sql:%s, row:%d col:%d data:%s == expect:%d" % + (sql, row, col, tdSql.queryResult[row][col], data)) + + return check_status + + def mycheckRows(self, sql, expectRows): + check_status = True + if len(tdSql.queryResult) == expectRows: + tdLog.info("sql:%s, queryRows:%d == expect:%d" % (sql, len(tdSql.queryResult), expectRows)) + return True + else: + caller = inspect.getframeinfo(inspect.stack()[1][0]) + args = (caller.filename, caller.lineno, sql, len(tdSql.queryResult), expectRows) + tdLog.info("%s(%d) failed: sql:%s, queryRows:%d != expect:%d" % args) + check_status = False + return check_status + + def stop_All(self): + + tdDnodes = cluster.dnodes + # newTdSql=tdCom.newTdSql() + # ==== stop all dnode ===== + for k ,v in self.dnode_list.items(): + dnode_id = v[0] + tdDnodes[dnode_id-1].stoptaosd() + # self.wait_stop_dnode_OK(newTdSql) + + def start_All(self): + tdDnodes = cluster.dnodes + # newTdSql=tdCom.newTdSql() + for k ,v in self.dnode_list.items(): + dnode_id = v[0] + start = time.time() + tdDnodes[dnode_id-1].starttaosd() + # self.wait_start_dnode_OK(newTdSql) + end = time.time() + time_cost = int(end -start) + if time_cost > self.max_restart_time: + tdLog.exit(" ==== restart dnode {} cost too much time , please check ====".format(self.stop_dnode_id)) + + + + def stop_All_dnodes_check_datas(self): + + # stop follower and insert datas , update tables and create new stables + self.create_database(dbname = self.db_name ,replica_num= self.replica , vgroup_nums= 1) + self.create_stable_insert_datas(dbname = self.db_name , stablename = "stb1" , tb_nums= self.tb_nums ,row_nums= self.row_nums ) + self.check_insert_rows(self.db_name ,'stb1' ,tb_nums = self.tb_nums , row_nums= self.row_nums ,append_rows=0) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + for i in range(self.loop_restart_times): + # begin to stop All taosd + self.stop_All() + # begin to start All taosd + self.start_All() + + tdLog.debug(" ==== cluster has restart , this is {}_th restart cluster ==== ".format(i)) + + self.check_insert_rows(self.db_name ,'stb1' ,tb_nums = self.tb_nums , row_nums= self.row_nums ,append_rows=0) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + self.append_rows_of_exists_tables(self.db_name ,"stb1" , "sub_stb1_0" , 100 ) + os.system(" taos -s ' select count(*) from {}.{} ;'".format(self.db_name ,'stb1' )) + + + def run(self): + + # basic insert and check of cluster + self.check_setup_cluster_status() + self.stop_All_dnodes_check_datas() + + + + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py index ce37f09d7d..c608c93f5e 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py @@ -220,25 +220,43 @@ class TDTestCase: status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) tdLog.debug(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + + + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='follower': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): + def wait_stop_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -258,10 +276,11 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {} ====".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): + def wait_start_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -369,12 +388,15 @@ class TDTestCase: def sync_run_case(self): # stop follower and insert datas , update tables and create new stables tdDnodes=cluster.dnodes + newTdSql = tdCom.newTdSql() for loop in range(self.loop_restart_times): db_name = "sync_db_{}".format(loop) stablename = 'stable_{}'.format(loop) self.create_database(dbname = db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = db_name , stablename = stablename , tb_nums= 10 ,row_nums= 10 ) - self.stop_dnode_id = self._get_stop_dnode_id(db_name) + + self.stop_dnode_id = self._get_stop_dnode_id(db_name , "follower") + # check rows of datas @@ -384,7 +406,9 @@ class TDTestCase: start = time.time() tdDnodes[self.stop_dnode_id-1].stoptaosd() - self.wait_stop_dnode_OK() + + self.wait_stop_dnode_OK(newTdSql) + # append rows of stablename when dnode stop @@ -402,7 +426,7 @@ class TDTestCase: # begin start dnode tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end -start) if time_cost > self.max_restart_time: @@ -419,7 +443,7 @@ class TDTestCase: def _restart_dnode_of_db_unsync(dbname): start = time.time() tdDnodes=cluster.dnodes - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + self.stop_dnode_id = self._get_stop_dnode_id(dbname ,"follower") # begin restart dnode tdDnodes[self.stop_dnode_id-1].stoptaosd() self.wait_stop_dnode_OK() diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py index 8869b5d3a9..25c26c7288 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py @@ -220,26 +220,42 @@ class TDTestCase: status_OK = self.mycheckRows("select distinct tbname from {}.{}".format(dbname,stablename) ,tb_nums) tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 + + def _get_stop_dnode_id(self,dbname ,dnode_role): - def _get_stop_dnode_id(self,dbname): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='follower': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): - + def wait_stop_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -259,10 +275,10 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): + def wait_start_dnode_OK(self ,newTdSql): def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -370,13 +386,16 @@ class TDTestCase: def sync_run_case(self): # stop follower and insert datas , update tables and create new stables tdDnodes=cluster.dnodes + newTdSql=tdCom.newTdSql() for loop in range(self.loop_restart_times): db_name = "sync_db_{}".format(loop) stablename = 'stable_{}'.format(loop) self.create_database(dbname = db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = db_name , stablename = stablename , tb_nums= 10 ,row_nums= 10 ) - self.stop_dnode_id = self._get_stop_dnode_id(db_name) + self.stop_dnode_id = self._get_stop_dnode_id(db_name ,"follower") + # print("dnode_id:" , self.stop_dnode_id ) + # check rows of datas self.check_insert_rows(db_name ,stablename ,tb_nums=10 , row_nums= 10 ,append_rows=0) @@ -384,8 +403,9 @@ class TDTestCase: # begin stop dnode start = time.time() tdDnodes[self.stop_dnode_id-1].stoptaosd() + + self.wait_stop_dnode_OK(newTdSql) - self.wait_stop_dnode_OK() # append rows of stablename when dnode stop @@ -403,7 +423,7 @@ class TDTestCase: # begin start dnode tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end -start) if time_cost > self.max_restart_time: @@ -418,14 +438,16 @@ class TDTestCase: def unsync_run_case(self): def _restart_dnode_of_db_unsync(dbname): + newTdSql=tdCom.newTdSql() start = time.time() tdDnodes=cluster.dnodes - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + self.stop_dnode_id = self._get_stop_dnode_id(dbname,"follower") + # print("dnode_id:" , self.stop_dnode_id ) # begin restart dnode tdDnodes[self.stop_dnode_id-1].stoptaosd() - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py index 0241cb5e32..edff274793 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py @@ -221,25 +221,41 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='follower': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): + def wait_stop_dnode_OK(self , newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -259,10 +275,11 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): + def wait_start_dnode_OK(self , newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -370,12 +387,15 @@ class TDTestCase: def sync_run_case(self): # stop follower and insert datas , update tables and create new stables tdDnodes=cluster.dnodes + newTdSql=tdCom.newTdSql() for loop in range(self.loop_restart_times): db_name = "sync_db_{}".format(loop) stablename = 'stable_{}'.format(loop) self.create_database(dbname = db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = db_name , stablename = stablename , tb_nums= 10 ,row_nums= 10 ) - self.stop_dnode_id = self._get_stop_dnode_id(db_name) + + self.stop_dnode_id = self._get_stop_dnode_id(db_name,"follower") + # check rows of datas @@ -384,8 +404,8 @@ class TDTestCase: # begin stop dnode start = time.time() tdDnodes[self.stop_dnode_id-1].forcestop() - - self.wait_stop_dnode_OK() + + self.wait_stop_dnode_OK(newTdSql) # append rows of stablename when dnode stop @@ -403,7 +423,7 @@ class TDTestCase: # begin start dnode tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end -start) if time_cost > self.max_restart_time: @@ -420,17 +440,21 @@ class TDTestCase: def _restart_dnode_of_db_unsync(dbname): start = time.time() tdDnodes=cluster.dnodes - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + newTdSql=tdCom.newTdSql() + self.stop_dnode_id = self._get_stop_dnode_id(dbname,"follower") + # print(self.stop_dnode_id) while not self.stop_dnode_id: + # print(self.stop_dnode_id) time.sleep(0.5) - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + self.stop_dnode_id = self._get_stop_dnode_id(dbname,"follower") # begin restart dnode # force stop taosd by kill -9 self.force_stop_dnode(self.stop_dnode_id) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) + os.system(" taos -s 'show dnodes;' ") tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py index df865e51fc..76bb2cc996 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py @@ -177,26 +177,40 @@ class TDTestCase: continue - def _get_stop_dnode_id(self,dbname): - newTdSql=tdCom.newTdSql() - newTdSql.query("show {}.vgroups".format(dbname)) - vgroup_infos = newTdSql.queryResult + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='leader': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): - + def wait_stop_dnode_OK(self , newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -216,10 +230,10 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): - + def wait_start_dnode_OK(self,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -278,6 +292,9 @@ class TDTestCase: os.system(" {} -f {} >>/dev/null 2>&1 ".format(benchmark_build_path , json_file)) def stop_leader_when_Benchmark_inserts(self,dbname , total_rows , json_file ): + + newTdSql=tdCom.newTdSql() + # stop follower and insert datas , update tables and create new stables tdDnodes=cluster.dnodes tdSql.execute(" drop database if exists {} ".format(dbname)) @@ -316,7 +333,7 @@ class TDTestCase: tdLog.debug(" === database {} has write {} rows at least ====".format(dbname,total_rows/10)) - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + self.stop_dnode_id = self._get_stop_dnode_id(dbname ,"leader") # prepare stop leader of database before_leader_infos = self.get_leader_infos(dbname) @@ -335,9 +352,9 @@ class TDTestCase: tdLog.debug(" ==== revote leader of database {} cost time {} ====".format(dbname , time_cost)) self.current_thread.join() - tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) + tdSql.query(" select count(*) from {}.{} ".format(dbname,"stb1")) tdLog.debug(" ==== expected insert {} rows of database {} , really is {}".format(total_rows, dbname , tdSql.queryResult[0][0])) diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py index 8d6f026165..6abe700bd6 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py @@ -33,10 +33,10 @@ class TDTestCase: self.tb_nums = 10 self.row_nums = 100 self.stop_dnode_id = None - self.loop_restart_times = 10 + self.loop_restart_times = 5 self.current_thread = None self.max_restart_time = 5 - + self.try_check_times = 10 def getBuildPath(self): selfPath = os.path.dirname(os.path.realpath(__file__)) if ("community" in selfPath): @@ -304,26 +304,40 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): - newTdSql=tdCom.newTdSql() - newTdSql.query("show {}.vgroups".format(dbname)) - vgroup_infos = newTdSql.queryResult + def _get_stop_dnode_id(self,dbname ,dnode_role): + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='leader': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): + def wait_stop_dnode_OK(self ,newTdSql): def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -343,10 +357,11 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): + def wait_start_dnode_OK(self,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -421,13 +436,15 @@ class TDTestCase: def sync_run_case(self): # stop follower and insert datas , update tables and create new stables tdDnodes=cluster.dnodes + newTdSql=tdCom.newTdSql() for loop in range(self.loop_restart_times): db_name = "sync_db_{}".format(loop) stablename = 'stable_{}'.format(loop) self.create_database(dbname = db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = db_name , stablename = stablename , tb_nums= 10 ,row_nums= 10 ) - self.stop_dnode_id = self._get_stop_dnode_id(db_name) + self.stop_dnode_id = self._get_stop_dnode_id(db_name ,"leader") + # check rows of datas self.check_insert_rows(db_name ,stablename ,tb_nums=10 , row_nums= 10 ,append_rows=0) @@ -439,7 +456,7 @@ class TDTestCase: # force stop taosd by kill -9 self.force_stop_dnode(self.stop_dnode_id) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) # vote leaders check @@ -451,6 +468,7 @@ class TDTestCase: # append rows of stablename when dnode stop make sure revote leaders while not revote_status: + after_leader_infos = self.get_leader_infos(db_name) revote_status = self.check_revote_leader_success(db_name ,before_leader_infos , after_leader_infos) @@ -473,7 +491,7 @@ class TDTestCase: # begin start dnode start = time.time() tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end -start) if time_cost > self.max_restart_time: @@ -490,14 +508,15 @@ class TDTestCase: def _restart_dnode_of_db_unsync(dbname): tdDnodes=cluster.dnodes - self.stop_dnode_id = self._get_stop_dnode_id(dbname) + newTdSql=tdCom.newTdSql() + self.stop_dnode_id = self._get_stop_dnode_id(dbname ,"leader" ) # begin restart dnode # force stop taosd by kill -9 # get leader info before stop before_leader_infos = self.get_leader_infos(db_name) self.force_stop_dnode(self.stop_dnode_id) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) # check revote leader when restart servers # get leader info after stop @@ -529,7 +548,7 @@ class TDTestCase: tdDnodes[self.stop_dnode_id-1].starttaosd() start = time.time() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_mnode3_insertdatas_querys.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_mnode3_insertdatas_querys.py new file mode 100644 index 0000000000..ed20a51595 --- /dev/null +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_mnode3_insertdatas_querys.py @@ -0,0 +1,218 @@ +# author : wenzhouwww +from ssl import ALERT_DESCRIPTION_CERTIFICATE_UNOBTAINABLE +import taos +import sys +import time +import os + +from util.log import * +from util.sql import * +from util.cases import * +from util.dnodes import TDDnodes +from util.dnodes import TDDnode +from util.cluster import * + +import time +import socket +import subprocess ,threading +sys.path.append(os.path.dirname(__file__)) + +class TDTestCase: + def init(self,conn ,logSql): + tdLog.debug(f"start to excute {__file__}") + tdSql.init(conn.cursor()) + self.host = socket.gethostname() + self.mnode_list = {} + self.dnode_list = {} + self.ts = 1483200000000 + self.db_name ='testdb' + self.replica = 3 + self.vgroups = 1 + self.tb_nums = 10 + self.row_nums = 1000 + self.query_times = 100 + + def getBuildPath(self): + selfPath = os.path.dirname(os.path.realpath(__file__)) + if ("community" in selfPath): + projPath = selfPath[:selfPath.find("community")] + else: + projPath = selfPath[:selfPath.find("tests")] + + for root, dirs, files in os.walk(projPath): + if ("taosd" in files): + rootRealPath = os.path.dirname(os.path.realpath(root)) + if ("packaging" not in rootRealPath): + buildPath = root[:len(root) - len("/build/bin")] + break + return buildPath + + def check_setup_cluster_status(self): + tdSql.query("show mnodes") + for mnode in tdSql.queryResult: + name = mnode[1] + info = mnode + self.mnode_list[name] = info + + tdSql.query("show dnodes") + for dnode in tdSql.queryResult: + name = dnode[1] + info = dnode + self.dnode_list[name] = info + + count = 0 + is_leader = False + mnode_name = '' + for k,v in self.mnode_list.items(): + count +=1 + # only for 1 mnode + mnode_name = k + + if v[2] =='leader': + is_leader=True + + if count==1 and is_leader: + tdLog.notice("===== depoly cluster success with 1 mnode as leader =====") + else: + tdLog.exit("===== depoly cluster fail with 1 mnode as leader =====") + + for k ,v in self.dnode_list.items(): + if k == mnode_name: + if v[3]==0: + tdLog.notice("===== depoly cluster mnode only success at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + tdLog.exit("===== depoly cluster mnode only fail at {} , support_vnodes is {} ".format(mnode_name,v[3])) + else: + continue + + def create_db_check_vgroups(self): + + tdSql.execute("drop database if exists test") + tdSql.execute("create database if not exists test replica 1 duration 300") + tdSql.execute("use test") + tdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + tdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(16),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(5): + tdSql.execute("create table sub_tb_{} using stb1 tags({})".format(i,i)) + tdSql.query("show stables") + tdSql.checkRows(1) + tdSql.query("show tables") + tdSql.checkRows(6) + + tdSql.query("show test.vgroups;") + vgroups_infos = {} # key is id: value is info list + for vgroup_info in tdSql.queryResult: + vgroup_id = vgroup_info[0] + tmp_list = [] + for role in vgroup_info[3:-4]: + if role in ['leader','follower']: + tmp_list.append(role) + vgroups_infos[vgroup_id]=tmp_list + + for k , v in vgroups_infos.items(): + if len(v) ==1 and v[0]=="leader": + tdLog.notice(" === create database replica only 1 role leader check success of vgroup_id {} ======".format(k)) + else: + tdLog.exit(" === create database replica only 1 role leader check fail of vgroup_id {} ======".format(k)) + + def create_db_replica_3_insertdatas(self, dbname, replica_num ,vgroup_nums ,tb_nums , row_nums ): + newTdSql=tdCom.newTdSql() + drop_db_sql = "drop database if exists {}".format(dbname) + create_db_sql = "create database {} replica {} vgroups {}".format(dbname,replica_num,vgroup_nums) + + tdLog.notice(" ==== create database {} and insert rows begin =====".format(dbname)) + newTdSql.execute(drop_db_sql) + newTdSql.execute(create_db_sql) + newTdSql.execute("use {}".format(dbname)) + newTdSql.execute( + '''create table stb1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + tags (t1 int) + ''' + ) + newTdSql.execute( + ''' + create table t1 + (ts timestamp, c1 int, c2 bigint, c3 smallint, c4 tinyint, c5 float, c6 double, c7 bool, c8 binary(32),c9 nchar(32), c10 timestamp) + ''' + ) + + for i in range(tb_nums): + sub_tbname = "sub_tb_{}".format(i) + newTdSql.execute("create table {} using stb1 tags({})".format(sub_tbname,i)) + # insert datas about new database + + for row_num in range(row_nums): + ts = self.ts + 1000*row_num + newTdSql.execute(f"insert into {sub_tbname} values ({ts}, {row_num} ,{row_num}, 10 ,1 ,{row_num} ,{row_num},true,'bin_{row_num}','nchar_{row_num}',now) ") + + tdLog.notice(" ==== create database {} and insert rows execute end =====".format(dbname)) + + + + def check_insert_status(self,newTdSql , dbname, tb_nums , row_nums): + # newTdSql=tdCom.newTdSql() + newTdSql.execute("use {}".format(dbname)) + newTdSql.query("select count(*) from {}.{}".format(dbname,'stb1')) + # tdSql.checkData(0 , 0 , tb_nums*row_nums) + newTdSql.query("select distinct tbname from {}.{}".format(dbname,'stb1')) + # tdSql.checkRows(tb_nums) + + def loop_query_constantly(self, times , db_name, tb_nums ,row_nums): + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === query is going ,this is {}_th query === ".format(loop_time)) + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + def loop_create_databases(self, times , tb_nums , row_nums): + newTdSql=tdCom.newTdSql() + for loop_time in range(times): + tdLog.debug(" === create database and insert datas is going ,this is {}_th create === ".format(loop_time)) + db_name = 'loop_db_{}'.format(loop_time) + self.create_db_replica_3_insertdatas(db_name , self.replica , self.vgroups , tb_nums , row_nums) + self.check_insert_status( newTdSql ,db_name, tb_nums , row_nums) + + def run(self): + self.check_setup_cluster_status() + self.create_db_check_vgroups() + + # create mnode + tdSql.execute("create mnode on dnode 2 ") + tdSql.execute("create mnode on dnode 3 ") + os.system("taos -s 'show mnodes;'") + # start writing constantly + writing = threading.Thread(target = self.create_db_replica_3_insertdatas, args=(self.db_name , self.replica , self.vgroups , self.tb_nums , self.row_nums)) + writing.start() + tdSql.query(" show {}.stables ".format(self.db_name)) + while not tdSql.queryResult: + print(tdSql.queryResult) + time.sleep(0.1) + tdSql.query(" show {}.stables ".format(self.db_name)) + + reading = threading.Thread(target = self.loop_query_constantly, args=(self.query_times,self.db_name , self.tb_nums , self.row_nums)) + reading.start() + + create_db = threading.Thread(target = self.loop_create_databases, args=(5, 10 , 10)) + create_db.start() + + writing.join() + reading.join() + create_db.join() + + def stop(self): + tdSql.close() + tdLog.success(f"{__file__} successfully executed") + +tdCases.addLinux(__file__, TDTestCase()) +tdCases.addWindows(__file__, TDTestCase()) \ No newline at end of file diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py index 4b404a5906..d60817a2b4 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py @@ -182,25 +182,40 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='follower': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): - + def wait_stop_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -220,10 +235,10 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): - + def wait_start_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -372,22 +387,24 @@ class TDTestCase: def stop_follower_when_query_going(self): tdDnodes = cluster.dnodes + newTdSql=tdCom.newTdSql() self.create_database(dbname = self.db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = self.db_name , stablename = "stb1" , tb_nums= self.tb_nums ,row_nums= self.row_nums) - # let query task start - self.thread_list = self.multi_thread_query_task(10 ,self.db_name ,'stb1' ) + # let query task start + self.thread_list = self.multi_thread_query_task(5 ,self.db_name ,'stb1' ) + # force stop follower for loop in range(self.loop_restart_times): tdLog.debug(" ==== this is {}_th restart follower of database {} ==== ".format(loop ,self.db_name)) - self.stop_dnode_id = self._get_stop_dnode_id(self.db_name) + self.stop_dnode_id = self._get_stop_dnode_id(self.db_name,"follower" ) tdDnodes[self.stop_dnode_id-1].stoptaosd() - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) start = time.time() tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) @@ -406,8 +423,6 @@ class TDTestCase: - - def stop(self): tdSql.close() tdLog.success(f"{__file__} successfully executed") diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py index 8310522bd9..da9dc115eb 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py @@ -182,25 +182,40 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='follower': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): + def wait_stop_dnode_OK(self ,newTdSql): def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -220,10 +235,10 @@ class TDTestCase: # tdLog.notice("==== stop dnode has not been stopped , endpoint is {}".format(self.stop_dnode)) tdLog.notice("==== stop_dnode has stopped , id is {}".format(self.stop_dnode_id)) - def wait_start_dnode_OK(self): - + def wait_start_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -379,15 +394,16 @@ class TDTestCase: self.thread_list = self.multi_thread_query_task(10 ,self.db_name ,'stb1' ) # force stop follower + newTdSql=tdCom.newTdSql() for loop in range(self.loop_restart_times): tdLog.debug(" ==== this is {}_th restart follower of database {} ==== ".format(loop ,self.db_name)) - self.stop_dnode_id = self._get_stop_dnode_id(self.db_name) + self.stop_dnode_id = self._get_stop_dnode_id(self.db_name,"follower" ) self.force_stop_dnode(self.stop_dnode_id) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) start = time.time() tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) @@ -405,9 +421,6 @@ class TDTestCase: self.stop_follower_when_query_going() - - - def stop(self): tdSql.close() tdLog.success(f"{__file__} successfully executed") diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py index 752a347d83..561159af89 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py @@ -182,25 +182,40 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='leader': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): - + def wait_stop_dnode_OK(self,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -242,10 +257,11 @@ class TDTestCase: break return check_status - def wait_start_dnode_OK(self): + def wait_start_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -349,11 +365,11 @@ class TDTestCase: tdLog.info("%s(%d) failed: sql:%s, queryRows:%d != expect:%d" % args) check_status = False return check_status - - - def get_leader_infos(self ,dbname): - - newTdSql=tdCom.newTdSql() + + + def get_leader_infos(self , newTdSql ,dbname): + + newTdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = newTdSql.queryResult @@ -409,39 +425,39 @@ class TDTestCase: self.create_database(dbname = self.db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = self.db_name , stablename = "stb1" , tb_nums= self.tb_nums ,row_nums= self.row_nums) - # let query task start - self.thread_list = self.multi_thread_query_task(10 ,self.db_name ,'stb1' ) + # let query task start + self.thread_list = self.multi_thread_query_task(2 ,self.db_name ,'stb1' ) + newTdSql=tdCom.newTdSql() # force stop follower for loop in range(self.loop_restart_times): tdLog.debug(" ==== this is {}_th restart follower of database {} ==== ".format(loop ,self.db_name)) - # get leader info before stop - before_leader_infos = self.get_leader_infos(self.db_name) + # get leader info before stop + before_leader_infos = self.get_leader_infos(newTdSql , self.db_name) - self.stop_dnode_id = self._get_stop_dnode_id(self.db_name) + self.stop_dnode_id = self._get_stop_dnode_id(self.db_name,"leader") tdDnodes[self.stop_dnode_id-1].stoptaosd() - start = time.time() - # get leader info after stop - after_leader_infos = self.get_leader_infos(self.db_name) - + # get leader info after stop + after_leader_infos = self.get_leader_infos(newTdSql , self.db_name) + revote_status = self.check_revote_leader_success(self.db_name ,before_leader_infos , after_leader_infos) while not revote_status: - after_leader_infos = self.get_leader_infos(self.db_name) + after_leader_infos = self.get_leader_infos(newTdSql , self.db_name) revote_status = self.check_revote_leader_success(self.db_name ,before_leader_infos , after_leader_infos) end = time.time() time_cost = end - start tdLog.debug(" ==== revote leader of database {} cost time {} ====".format(self.db_name , time_cost)) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) start = time.time() tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) @@ -459,9 +475,6 @@ class TDTestCase: self.stop_follower_when_query_going() - - - def stop(self): tdSql.close() tdLog.success(f"{__file__} successfully executed") diff --git a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py index 9f64faa446..fb0ddd5435 100644 --- a/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py +++ b/tests/system-test/6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py @@ -182,25 +182,40 @@ class TDTestCase: tdLog.notice(" ==== check insert tbnames first failed , this is {}_th retry check tbnames of database {}".format(count , dbname)) count += 1 - def _get_stop_dnode_id(self,dbname): + def _get_stop_dnode_id(self,dbname ,dnode_role): tdSql.query("show {}.vgroups".format(dbname)) vgroup_infos = tdSql.queryResult + status = False + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + while status!=True : + time.sleep(0.1) + tdSql.query("show {}.vgroups".format(dbname)) + vgroup_infos = tdSql.queryResult + for vgroup_info in vgroup_infos: + if "error" not in vgroup_info: + status = True + else: + status = False + # print(status) for vgroup_info in vgroup_infos: leader_infos = vgroup_info[3:-4] # print(vgroup_info) for ind ,role in enumerate(leader_infos): - if role =='leader': + if role == dnode_role: # print(ind,leader_infos) self.stop_dnode_id = leader_infos[ind-1] break - return self.stop_dnode_id - def wait_stop_dnode_OK(self): + def wait_stop_dnode_OK(self ,newTdSql): def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") @@ -242,10 +257,10 @@ class TDTestCase: break return check_status - def wait_start_dnode_OK(self): - + def wait_start_dnode_OK(self ,newTdSql): + def _get_status(): - newTdSql=tdCom.newTdSql() + # newTdSql=tdCom.newTdSql() status = "" newTdSql.query("show dnodes") dnode_infos = newTdSql.queryResult @@ -406,11 +421,13 @@ class TDTestCase: def stop_follower_when_query_going(self): tdDnodes = cluster.dnodes + newTdSql=tdCom.newTdSql() self.create_database(dbname = self.db_name ,replica_num= self.replica , vgroup_nums= 1) self.create_stable_insert_datas(dbname = self.db_name , stablename = "stb1" , tb_nums= self.tb_nums ,row_nums= self.row_nums) - # let query task start - self.thread_list = self.multi_thread_query_task(10 ,self.db_name ,'stb1' ) + # let query task start + self.thread_list = self.multi_thread_query_task(5 ,self.db_name ,'stb1' ) + # force stop follower for loop in range(self.loop_restart_times): @@ -419,7 +436,7 @@ class TDTestCase: # get leader info before stop before_leader_infos = self.get_leader_infos(self.db_name) - self.stop_dnode_id = self._get_stop_dnode_id(self.db_name) + self.stop_dnode_id = self._get_stop_dnode_id(self.db_name ,"leader") self.force_stop_dnode(self.stop_dnode_id) @@ -437,11 +454,11 @@ class TDTestCase: time_cost = end - start tdLog.debug(" ==== revote leader of database {} cost time {} ====".format(self.db_name , time_cost)) - self.wait_stop_dnode_OK() + self.wait_stop_dnode_OK(newTdSql) start = time.time() tdDnodes[self.stop_dnode_id-1].starttaosd() - self.wait_start_dnode_OK() + self.wait_start_dnode_OK(newTdSql) end = time.time() time_cost = int(end-start) diff --git a/tests/system-test/clusterCase.sh b/tests/system-test/clusterCase.sh index cfc44f7f95..a4bec11d03 100755 --- a/tests/system-test/clusterCase.sh +++ b/tests/system-test/clusterCase.sh @@ -19,4 +19,28 @@ python3 ./test.py -f 6-cluster/5dnode3mnodeSep1VnodeStopMnodeCreateStb.py -N 5 # python3 ./test.py -f 6-cluster/5dnode3mnodeRestartDnodeInsertData.py -N 5 -M 3 python3 ./test.py -f 6-cluster/5dnode3mnodeAdd1Ddnoe.py -N 6 -M 3 -C 5 - +# test case of vnode +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_createDb_replica1.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_vgroups.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_vgroups_stopOne.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py  -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py  -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas_querys.py -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys.py -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_follower.py -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_leader.py -N 4 -M 1 +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_all_vnode.py   -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_all_dnodes.py -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_force_stop_all_dnodes.py -N 4 -M 1  +python3 ./test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_mnode3_insertdatas_querys.py -N 4 -M 1 + diff --git a/tests/system-test/fulltest.sh b/tests/system-test/fulltest.sh index 7a3ad1070c..75e3664bbd 100755 --- a/tests/system-test/fulltest.sh +++ b/tests/system-test/fulltest.sh @@ -193,6 +193,30 @@ python3 ./test.py -f 6-cluster/5dnode3mnodeRecreateMnode.py -N 5 -M 3 python3 ./test.py -f 6-cluster/5dnode3mnodeStopFollowerLeader.py -N 5 -M 3 python3 ./test.py -f 6-cluster/5dnode3mnodeStop2Follower.py -N 5 -M 3 +# vnode case +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_createDb_replica1.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica1_insertdatas_querys.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_force_stop_all_dnodes.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_all_vnode.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_follower.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys_loop_restart_leader.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_querys.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_all_dnodes.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_sync.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync_force_stop.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_follower_unsync.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader_forece_stop.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_insertdatas_stop_leader.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_mnode3_insertdatas_querys.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower_force_stop.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_follower.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader_force_stop.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_querydatas_stop_leader.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_vgroups.py -N 4 -M 1 +# python3 test.py -f 6-cluster/vnode/4dnode1mnode_basic_replica3_vgroups_stopOne.py -N 4 -M 1 + python3 ./test.py -f 7-tmq/dropDbR3ConflictTransaction.py -N 3 python3 ./test.py -f 7-tmq/basic5.py