Elgato_dark/TDengine
1
0
Fork 0
mirror of https://github.com/taosdata/TDengine synced 2026-05-24 10:09:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 5
TDengine/test/cases/11-Functions/02-Aggregate/test_fun_agg_spread.py

655 lines
No EOL
28 KiB
Python
Raw Blame History

import datetime
import numpy as np
import random
from new_test_framework.utils import tdLog, tdSql, sc, clusterComCheck, tdDnodes
# ------ global test_spread.py ------
PRIMARY_COL = "ts"
INT_COL = "c1"
BINT_COL = "c2"
SINT_COL = "c3"
TINT_COL = "c4"
FLOAT_COL = "c5"
DOUBLE_COL = "c6"
BOOL_COL = "c7"
BINARY_COL = "c8"
NCHAR_COL = "c9"
TS_COL = "c10"
NUM_COL = [ INT_COL, BINT_COL, SINT_COL, TINT_COL, FLOAT_COL, DOUBLE_COL, ]
CHAR_COL = [ BINARY_COL, NCHAR_COL, ]
BOOLEAN_COL = [ BOOL_COL, ]
TS_TYPE_COL = [ TS_COL, ]
ALL_COL = [ INT_COL, BINT_COL, SINT_COL, TINT_COL, FLOAT_COL, DOUBLE_COL, BOOL_COL, BINARY_COL, NCHAR_COL, TS_COL ]
DBNAME = "db"
class TestFunSpread:
# distribute
updatecfgDict = {"maxTablesPerVnode":2 ,"minTablesPerVnode":2,"tableIncStepPerVnode":2 }
def setup_class(cls):
tdLog.debug(f"start to execute {__file__}")
#
# ------------------ sim case ------------------
#
def do_sim_spread(self):
tdSql.execute(f"create database test KEEP 36500;")
tdSql.execute(f"use test;")
tdSql.execute(f"create table st(ts timestamp, f int) tags(t int);")
ms = 1712135244502
ms1 = ms + 1000
ms2 = ms + 2000
ms3 = ms + 3000
ms4 = ms + 4000
ms5 = ms + 5000
ms6 = ms + 6000
ms7 = ms + 7000
tdSql.execute(
f"insert into ct1 using st tags(1) values({ms} , 0)({ms1} , 1)({ms2} , 10)({ms3} , 11)"
)
tdSql.execute(f"insert into ct2 using st tags(2) values({ms2} , 2)({ms3} , 3)")
tdSql.execute(f"insert into ct3 using st tags(3) values({ms4} , 4)({ms5} , 5)")
tdSql.execute(f"insert into ct4 using st tags(4) values({ms6} , 6)({ms7} , 7)")
tdSql.query(f"select count(*), spread(ts) from st where tbname='ct1'")
tdLog.info(f"{tdSql.getData(0,0)}, {tdSql.getData(0,1)}")
tdSql.checkData(0, 0, 4)
tdSql.checkData(0, 1, 3000.000000000)
tdSql.execute(f"drop database test;")
print("\n")
print("do_sim_spread ......................... [passed]")
#
# ------------------ test_spread.py ------------------
#
def __query_condition(self,tbname):
query_condition = [f"cast({col} as bigint)" for col in ALL_COL]
for num_col in NUM_COL:
query_condition.extend(
(
f"{tbname}.{num_col}",
f"abs( {tbname}.{num_col} )",
f"acos( {tbname}.{num_col} )",
f"asin( {tbname}.{num_col} )",
f"atan( {tbname}.{num_col} )",
f"avg( {tbname}.{num_col} )",
f"ceil( {tbname}.{num_col} )",
f"cos( {tbname}.{num_col} )",
f"count( {tbname}.{num_col} )",
f"floor( {tbname}.{num_col} )",
f"log( {tbname}.{num_col}, {tbname}.{num_col})",
f"max( {tbname}.{num_col} )",
f"min( {tbname}.{num_col} )",
f"pow( {tbname}.{num_col}, 2)",
f"round( {tbname}.{num_col} )",
f"sum( {tbname}.{num_col} )",
f"sin( {tbname}.{num_col} )",
f"sqrt( {tbname}.{num_col} )",
f"tan( {tbname}.{num_col} )",
f"cast( {tbname}.{num_col} as timestamp)",
)
)
[ query_condition.append(f"{num_col} + {any_col}") for any_col in ALL_COL ]
for char_col in CHAR_COL:
query_condition.extend(
(
f"count({tbname}.{char_col})",
f"sum(cast({tbname}.{char_col}) as bigint)",
f"max(cast({tbname}.{char_col}) as bigint)",
f"min(cast({tbname}.{char_col}) as bigint)",
f"avg(cast({tbname}.{char_col}) as bigint)",
)
)
query_condition.extend(
(
1010.1,
)
)
return query_condition
def __join_condition(self, tb_list, filter=PRIMARY_COL, INNER=False):
table_reference = tb_list[0]
join_condition = table_reference
join = "inner join" if INNER else "join"
for i in range(len(tb_list[1:])):
join_condition += f" {join} {tb_list[i+1]} on {table_reference}.{filter}={tb_list[i+1]}.{filter}"
return join_condition
def __where_condition(self, col=None, tbname=None, query_conditon=None):
# tbname = tbname.split(".")[-1] if tbname else None
if query_conditon and isinstance(query_conditon, str):
if query_conditon.startswith("count"):
query_conditon = query_conditon[6:-1]
elif query_conditon.startswith("max"):
query_conditon = query_conditon[4:-1]
elif query_conditon.startswith("sum"):
query_conditon = query_conditon[4:-1]
elif query_conditon.startswith("min"):
query_conditon = query_conditon[4:-1]
if query_conditon:
return f" where {query_conditon} is not null"
if col in NUM_COL:
return f" where abs( {tbname}.{col} ) >= 0"
if col in CHAR_COL:
return f" where lower( {tbname}.{col} ) like 'bina%' or lower( {tbname}.{col} ) like '_cha%' "
if col in BOOLEAN_COL:
return f" where {tbname}.{col} in (false, true) "
if col in TS_TYPE_COL or col in PRIMARY_COL:
return f" where cast( {tbname}.{col} as binary(16) ) is not null "
return ""
def __group_condition(self, col, having = None):
if isinstance(col, str):
if col.startswith("count"):
col = col[6:-1]
elif col.startswith("max"):
col = col[4:-1]
elif col.startswith("sum"):
col = col[4:-1]
elif col.startswith("min"):
col = col[4:-1]
return f" group by {col} having {having}" if having else f" group by {col} "
def __single_sql(self, select_clause, from_clause, where_condition="", group_condition=""):
if isinstance(select_clause, str) and "on" not in from_clause and select_clause.split(".")[0] != from_clause.split(".")[0]:
return
return f"select spread({select_clause}) from {from_clause} {where_condition} {group_condition}"
@property
def __tb_list(self, dbname=DBNAME):
return [
f"{dbname}.ct1",
f"{dbname}.ct4",
f"{dbname}.t1",
f"{dbname}.ct2",
f"{dbname}.stb1",
]
def sql_list(self):
sqls = []
__no_join_tblist = self.__tb_list
for tb in __no_join_tblist:
tbname = tb.split(".")[-1]
select_claus_list = self.__query_condition(tbname)
for select_claus in select_claus_list:
group_claus = self.__group_condition(col=select_claus)
where_claus = self.__where_condition(query_conditon=select_claus)
having_claus = self.__group_condition(col=select_claus, having=f"{select_claus} is not null")
sqls.extend(
(
self.__single_sql(select_claus, tb, where_claus, having_claus),
self.__single_sql(select_claus, tb,),
self.__single_sql(select_claus, tb, where_condition=where_claus),
self.__single_sql(select_claus, tb, group_condition=group_claus),
)
)
# return filter(None, sqls)
return list(filter(None, sqls))
def spread_check(self):
sqls = self.sql_list()
tdLog.printNoPrefix("===step 1: curent case, must return query OK")
for i in range(len(sqls)):
tdLog.info(f"sql: {sqls[i]}")
tdSql.query(sqls[i])
def __test_current(self, dbname=DBNAME):
tdSql.query(f"select spread(ts) from {dbname}.ct1")
tdSql.checkRows(1)
tdSql.query(f"select spread(c1) from {dbname}.ct2")
tdSql.checkRows(1)
tdSql.query(f"select spread(c1) from {dbname}.ct4 group by c1")
tdSql.checkRows(self.rows + 3)
tdSql.query(f"select spread(c1) from {dbname}.ct4 group by c7")
tdSql.checkRows(3)
tdSql.query(f"select spread(ct2.c1) from {dbname}.ct4 ct4 join {dbname}.ct2 ct2 on ct4.ts=ct2.ts")
tdSql.checkRows(1)
self.spread_check()
def __test_error(self, dbname=DBNAME):
tdLog.printNoPrefix("===step 0: err case, must return err")
tdSql.error( f"select spread() from {dbname}.ct1" )
tdSql.error( f"select spread(1, 2) from {dbname}.ct2" )
tdSql.error( f"select spread({NUM_COL[0]}, {NUM_COL[1]}) from {dbname}.ct4" )
tdSql.error( f"select spread({BOOLEAN_COL[0]}) from {dbname}.t1" )
tdSql.error( f"select spread({CHAR_COL[0]}) from {dbname}.stb1" )
# tdSql.error( ''' select spread(['c1 + c1', 'c1 + c2', 'c1 + c3', 'c1 + c4', 'c1 + c5', 'c1 + c6', 'c1 + c7', 'c1 + c8', 'c1 + c9', 'c1 + c10'])
# from ct1
# where ['c1 + c1', 'c1 + c2', 'c1 + c3', 'c1 + c4', 'c1 + c5', 'c1 + c6', 'c1 + c7', 'c1 + c8', 'c1 + c9', 'c1 + c10'] is not null
# group by ['c1 + c1', 'c1 + c2', 'c1 + c3', 'c1 + c4', 'c1 + c5', 'c1 + c6', 'c1 + c7', 'c1 + c8', 'c1 + c9', 'c1 + c10']
# having ['c1 + c1', 'c1 + c2', 'c1 + c3', 'c1 + c4', 'c1 + c5', 'c1 + c6', 'c1 + c7', 'c1 + c8', 'c1 + c9', 'c1 + c10'] is not null ''' )
# tdSql.error( "select c1 from ct1 union select c1 from ct2 union select c1 from ct4 ")
def all_test(self, dbname=DBNAME):
self.__test_error(dbname)
self.__test_current(dbname)
def __create_tb(self, dbname=DBNAME):
tdLog.printNoPrefix("==========step1:create table")
create_stb_sql = f'''create table {dbname}.stb1(
ts timestamp, {INT_COL} int, {BINT_COL} bigint, {SINT_COL} smallint, {TINT_COL} tinyint,
{FLOAT_COL} float, {DOUBLE_COL} double, {BOOL_COL} bool,
{BINARY_COL} binary(16), {NCHAR_COL} nchar(32), {TS_COL} timestamp
) tags (t1 int)
'''
create_ntb_sql = f'''create table {dbname}.t1(
ts timestamp, {INT_COL} int, {BINT_COL} bigint, {SINT_COL} smallint, {TINT_COL} tinyint,
{FLOAT_COL} float, {DOUBLE_COL} double, {BOOL_COL} bool,
{BINARY_COL} binary(16), {NCHAR_COL} nchar(32), {TS_COL} timestamp
)
'''
tdSql.execute(create_stb_sql)
tdSql.execute(create_ntb_sql)
for i in range(4):
tdSql.execute(f'create table {dbname}.ct{i+1} using {dbname}.stb1 tags ( {i+1} )')
{ i % 32767 }, { i % 127}, { i * 1.11111 }, { i * 1000.1111 }, { i % 2}
def __insert_data(self, rows, dbname=DBNAME):
now_time = int(datetime.datetime.timestamp(datetime.datetime.now()) * 1000)
for i in range(rows):
tdSql.execute(
f"insert into {dbname}.ct1 values ( { now_time - i * 1000 }, {i}, {11111 * i}, {111 * i % 32767 }, {11 * i % 127}, {1.11*i}, {1100.0011*i}, {i%2}, 'binary{i}', 'nchar_测试_{i}', { now_time + 1 * i } )"
)
tdSql.execute(
f"insert into {dbname}.ct4 values ( { now_time - i * 7776000000 }, {i}, {11111 * i}, {111 * i % 32767 }, {11 * i % 127}, {1.11*i}, {1100.0011*i}, {i%2}, 'binary{i}', 'nchar_测试_{i}', { now_time + 1 * i } )"
)
tdSql.execute(
f"insert into {dbname}.ct2 values ( { now_time - i * 7776000000 }, {-i}, {-11111 * i}, {-111 * i % 32767 }, {-11 * i % 127}, {-1.11*i}, {-1100.0011*i}, {i%2}, 'binary{i}', 'nchar_测试_{i}', { now_time + 1 * i } )"
)
tdSql.execute(
f'''insert into {dbname}.ct1 values
( { now_time - rows * 5 }, 0, 0, 0, 0, 0, 0, 0, 'binary0', 'nchar_测试_0', { now_time + 8 } )
( { now_time + 10000 }, { rows }, -99999, -999, -99, -9.99, -99.99, 1, 'binary9', 'nchar_测试_9', { now_time + 9 } )
'''
)
tdSql.execute(
f'''insert into {dbname}.ct4 values
( { now_time - rows * 7776000000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time - rows * 3888000000 + 10800000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time + 7776000000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
(
{ now_time + 5184000000}, {pow(2,31)-pow(2,15)}, {pow(2,63)-pow(2,30)}, 32767, 127,
{ 3.3 * pow(10,38) }, { 1.3 * pow(10,308) }, { rows % 2 }, "binary_limit-1", "nchar_测试_limit-1", { now_time - 86400000}
)
(
{ now_time + 2592000000 }, {pow(2,31)-pow(2,16)}, {pow(2,63)-pow(2,31)}, 32766, 126,
{ 3.2 * pow(10,38) }, { 1.2 * pow(10,308) }, { (rows-1) % 2 }, "binary_limit-2", "nchar_测试_limit-2", { now_time - 172800000}
)
'''
)
tdSql.execute(
f'''insert into {dbname}.ct2 values
( { now_time - rows * 7776000000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time - rows * 3888000000 + 10800000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time + 7776000000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
(
{ now_time + 5184000000 }, { -1 * pow(2,31) + pow(2,15) }, { -1 * pow(2,63) + pow(2,30) }, -32766, -126,
{ -1 * 3.2 * pow(10,38) }, { -1.2 * pow(10,308) }, { rows % 2 }, "binary_limit-1", "nchar_测试_limit-1", { now_time - 86400000 }
)
(
{ now_time + 2592000000 }, { -1 * pow(2,31) + pow(2,16) }, { -1 * pow(2,63) + pow(2,31) }, -32767, -127,
{ - 3.3 * pow(10,38) }, { -1.3 * pow(10,308) }, { (rows-1) % 2 }, "binary_limit-2", "nchar_测试_limit-2", { now_time - 172800000 }
)
'''
)
for i in range(rows):
insert_data = f'''insert into {dbname}.t1 values
( { now_time - i * 3600000 }, {i}, {i * 11111}, { i % 32767 }, { i % 127}, { i * 1.11111 }, { i * 1000.1111 }, { i % 2},
"binary_{i}", "nchar_测试_{i}", { now_time - 1000 * i } )
'''
tdSql.execute(insert_data)
tdSql.execute(
f'''insert into {dbname}.t1 values
( { now_time + 10800000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time - (( rows // 2 ) * 60 + 30) * 60000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time - rows * 3600000 }, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( { now_time + 7200000 }, { pow(2,31) - pow(2,15) }, { pow(2,63) - pow(2,30) }, 32767, 127,
{ 3.3 * pow(10,38) }, { 1.3 * pow(10,308) }, { rows % 2 },
"binary_limit-1", "nchar_测试_limit-1", { now_time - 86400000 }
)
(
{ now_time + 3600000 } , { pow(2,31) - pow(2,16) }, { pow(2,63) - pow(2,31) }, 32766, 126,
{ 3.2 * pow(10,38) }, { 1.2 * pow(10,308) }, { (rows-1) % 2 },
"binary_limit-2", "nchar_测试_limit-2", { now_time - 172800000 }
)
'''
)
# do spread
def do_spread(self):
tdSql.prepare()
tdLog.printNoPrefix("==========step1:create table")
self.__create_tb()
tdLog.printNoPrefix("==========step2:insert data")
self.rows = 10
self.__insert_data(self.rows)
tdLog.printNoPrefix("==========step3:all check")
self.all_test()
tdDnodes.stop(1)
tdDnodes.start(1)
tdSql.execute("use db")
tdLog.printNoPrefix("==========step4:after wal, all check again ")
self.all_test()
print("do_spread ............................. [passed]")
#
# ------------------ test_distribute_spread.py ------------------
#
def check_spread_functions(self, tbname , col_name):
spread_sql = f"select spread({col_name}) from {tbname};"
same_sql = f"select max({col_name})-min({col_name}) from {tbname}"
tdSql.query(spread_sql)
spread_result = tdSql.queryResult
tdSql.query(same_sql)
same_result = tdSql.queryResult
if spread_result !=same_result:
tdLog.exit(f" max function work not as expected, sql : {spread_sql} ")
else:
tdLog.info(f" max function work as expected, sql : {spread_sql} ")
def prepare_datas_of_distribute(self, dbname="testdb"):
# prepate datas for 20 tables distributed at different vgroups
tdSql.execute(f"create database if not exists {dbname} keep 3650 duration 100 vgroups 5")
tdSql.execute(f" use {dbname}")
tdSql.execute(
f'''create table {dbname}.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 (t0 timestamp, t1 int, t2 bigint, t3 smallint, t4 tinyint, t5 float, t6 double, t7 bool, t8 binary(16),t9 nchar(32))
'''
)
tdSql.execute(
f'''
create table {dbname}.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(20):
tdSql.execute(f'create table {dbname}.ct{i+1} using {dbname}.stb1 tags ( now(), {1*i}, {11111*i}, {111*i}, {1*i}, {1.11*i}, {11.11*i}, {i%2}, "binary{i}", "nchar{i}" )')
for i in range(9):
tdSql.execute(
f"insert into {dbname}.ct1 values ( now()-{i*10}s, {1*i}, {11111*i}, {111*i}, {11*i}, {1.11*i}, {11.11*i}, {i%2}, 'binary{i}', 'nchar{i}', now()+{1*i}a )"
)
tdSql.execute(
f"insert into {dbname}.ct4 values ( now()-{i*90}d, {1*i}, {11111*i}, {111*i}, {11*i}, {1.11*i}, {11.11*i}, {i%2}, 'binary{i}', 'nchar{i}', now()+{1*i}a )"
)
for i in range(1,21):
if i ==1 or i == 4:
continue
else:
tbname = f"ct{i}"
for j in range(9):
tdSql.execute(
f"insert into {dbname}.{tbname} values ( now()-{(i+j)*10}s, {1*(j+i)}, {11111*(j+i)}, {111*(j+i)}, {11*(j)}, {1.11*(j+i)}, {11.11*(j+i)}, {(j+i)%2}, 'binary{j}', 'nchar{j}', now()+{1*j}a )"
)
tdSql.execute(f"insert into {dbname}.ct1 values (now()-45s, 0, 0, 0, 0, 0, 0, 0, 'binary0', 'nchar0', now()+8a )")
tdSql.execute(f"insert into {dbname}.ct1 values (now()+10s, 9, -99999, -999, -99, -9.99, -99.99, 1, 'binary9', 'nchar9', now()+9a )")
tdSql.execute(f"insert into {dbname}.ct1 values (now()+15s, 9, -99999, -999, -99, -9.99, NULL, 1, 'binary9', 'nchar9', now()+9a )")
tdSql.execute(f"insert into {dbname}.ct1 values (now()+20s, 9, -99999, -999, NULL, -9.99, -99.99, 1, 'binary9', 'nchar9', now()+9a )")
tdSql.execute(f"insert into {dbname}.ct4 values (now()-810d, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL ) ")
tdSql.execute(f"insert into {dbname}.ct4 values (now()-400d, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL ) ")
tdSql.execute(f"insert into {dbname}.ct4 values (now()+90d, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL ) ")
tdSql.execute(
f'''insert into {dbname}.t1 values
( '2020-04-21 01:01:01.000', NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( '2020-10-21 01:01:01.000', 1, 11111, 111, 11, 1.11, 11.11, 1, "binary1", "nchar1", now()+1a )
( '2020-12-31 01:01:01.000', 2, 22222, 222, 22, 2.22, 22.22, 0, "binary2", "nchar2", now()+2a )
( '2021-01-01 01:01:06.000', 3, 33333, 333, 33, 3.33, 33.33, 0, "binary3", "nchar3", now()+3a )
( '2021-05-07 01:01:10.000', 4, 44444, 444, 44, 4.44, 44.44, 1, "binary4", "nchar4", now()+4a )
( '2021-07-21 01:01:01.000', NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
( '2021-09-30 01:01:16.000', 5, 55555, 555, 55, 5.55, 55.55, 0, "binary5", "nchar5", now()+5a )
( '2022-02-01 01:01:20.000', 6, 66666, 666, 66, 6.66, 66.66, 1, "binary6", "nchar6", now()+6a )
( '2022-10-28 01:01:26.000', 7, 00000, 000, 00, 0.00, 00.00, 1, "binary7", "nchar7", "1970-01-01 08:00:00.000" )
( '2022-12-01 01:01:30.000', 8, -88888, -888, -88, -8.88, -88.88, 0, "binary8", "nchar8", "1969-01-01 01:00:00.000" )
( '2022-12-31 01:01:36.000', 9, -99999999999999999, -999, -99, -9.99, -999999999999999999999.99, 1, "binary9", "nchar9", "1900-01-01 00:00:00.000" )
( '2023-02-21 01:01:01.000', NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL )
'''
)
tdLog.info(f" prepare data for distributed_aggregate done! ")
def check_distribute_datas(self, dbname="testdb"):
# get vgroup_ids of all
tdSql.query(f"show {dbname}.vgroups ")
vgroups = tdSql.queryResult
vnode_tables={}
for vgroup_id in vgroups:
vnode_tables[vgroup_id[0]]=[]
# check sub_table of per vnode ,make sure sub_table has been distributed
tdSql.query(f"select * from information_schema.ins_tables where db_name = '{dbname}' and table_name like 'ct%'")
table_names = tdSql.queryResult
tablenames = []
for table_name in table_names:
vnode_tables[table_name[6]].append(table_name[0])
self.vnode_disbutes = vnode_tables
count = 0
for k ,v in vnode_tables.items():
if len(v)>=2:
count+=1
if count < 2:
tdLog.exit(f" the datas of all not satisfy sub_table has been distributed ")
def check_spread_distribute_diff_vnode(self,col_name, dbname="testdb"):
vgroup_ids = []
for k ,v in self.vnode_disbutes.items():
if len(v)>=2:
vgroup_ids.append(k)
distribute_tbnames = []
for vgroup_id in vgroup_ids:
vnode_tables = self.vnode_disbutes[vgroup_id]
distribute_tbnames.append(random.sample(vnode_tables,1)[0])
tbname_ins = ""
for tbname in distribute_tbnames:
tbname_ins += f"'{tbname}' ,"
tbname_filters = tbname_ins[:-1]
spread_sql = f"select spread({col_name}) from {dbname}.stb1 where tbname in ({tbname_filters})"
same_sql = f"select max({col_name}) - min({col_name}) from {dbname}.stb1 where tbname in ({tbname_filters})"
tdSql.query(spread_sql)
spread_result = tdSql.queryResult
tdSql.query(same_sql)
same_result = tdSql.queryResult
if spread_result !=same_result:
tdLog.exit(f" spread function work not as expected, sql : {spread_sql} ")
else:
tdLog.info(f" spread function work as expected, sql : {spread_sql} ")
def check_spread_status(self, dbname="testdb"):
# check max function work status
tdSql.query(f"show {dbname}.tables like 'ct%'")
table_names = tdSql.queryResult
tablenames = []
for table_name in table_names:
tablenames.append(f"{dbname}.{table_name[0]}")
tdSql.query(f"desc {dbname}.stb1")
col_names = tdSql.queryResult
colnames = []
for col_name in col_names:
if col_name[1] in ["INT" ,"BIGINT" ,"SMALLINT" ,"TINYINT" , "FLOAT" ,"DOUBLE"]:
colnames.append(col_name[0])
for tablename in tablenames:
for colname in colnames:
self.check_spread_functions(tablename,colname)
# check max function for different vnode
for colname in colnames:
if colname.startswith(f"c"):
self.check_spread_distribute_diff_vnode(colname)
def distribute_agg_query(self, dbname="testdb"):
# basic filter
tdSql.query(f"select spread(c1) from {dbname}.stb1 where c1 is null")
tdSql.checkRows(1)
tdSql.query(f"select spread(c1) from {dbname}.stb1 where t1=1")
tdSql.checkData(0,0,8.000000000)
tdSql.query(f"select spread(c1+c2) from {dbname}.stb1 where c1 =1 ")
tdSql.checkData(0,0,0.000000000)
tdSql.query(f"select spread(c1) from {dbname}.stb1 where tbname=\"ct2\"")
tdSql.checkData(0,0,8.000000000)
tdSql.query(f"select spread(c1) from {dbname}.stb1 partition by tbname")
tdSql.checkRows(20)
tdSql.query(f"select spread(c1) from {dbname}.stb1 where t1> 4 partition by tbname")
tdSql.checkRows(15)
# union all
tdSql.query(f"select spread(c1) from {dbname}.stb1 union all select max(c1)-min(c1) from {dbname}.stb1 ")
tdSql.checkRows(2)
tdSql.checkData(0,0,28.000000000)
# join
tdSql.execute(f" create database if not exists db ")
tdSql.execute(f" use db ")
tdSql.execute(f" create stable db.st (ts timestamp , c1 int ,c2 float) tags(t1 int) ")
tdSql.execute(f" create table db.tb1 using db.st tags(1) ")
tdSql.execute(f" create table db.tb2 using db.st tags(2) ")
for i in range(10):
ts = i*10 + self.ts
tdSql.execute(f" insert into db.tb1 values({ts},{i},{i}.0)")
tdSql.execute(f" insert into db.tb2 values({ts},{i},{i}.0)")
tdSql.query(f"select spread(tb1.c1), spread(tb2.c2) from db.tb1 tb1, db.tb2 tb2 where tb1.ts=tb2.ts")
tdSql.checkRows(1)
tdSql.checkData(0,0,9.000000000)
tdSql.checkData(0,0,9.00000)
# group by
tdSql.execute(f" use {dbname} ")
tdSql.query(f" select max(c1),c1 from {dbname}.stb1 group by t1 ")
tdSql.checkRows(20)
tdSql.query(f" select max(c1),c1 from {dbname}.stb1 group by c1 ")
tdSql.checkRows(30)
tdSql.query(f" select max(c1),c2 from {dbname}.stb1 group by c2 ")
tdSql.checkRows(31)
# partition by tbname or partition by tag
tdSql.query(f"select spread(c1) from {dbname}.stb1 partition by tbname")
query_data = tdSql.queryResult
# nest query for support max
tdSql.query(f"select spread(c2+2)+1 from (select max(c1) c2 from {dbname}.stb1)")
tdSql.checkData(0,0,1.000000000)
tdSql.query(f"select spread(c1+2)+1 as c2 from (select ts ,c1 ,c2 from {dbname}.stb1)")
tdSql.checkData(0,0,29.000000000)
tdSql.query(f"select spread(a+2)+1 as c2 from (select ts ,abs(c1) a ,c2 from {dbname}.stb1)")
tdSql.checkData(0,0,29.000000000)
# mixup with other functions
tdSql.query(f"select max(c1),count(c1),last(c2,c3),spread(c1) from {dbname}.stb1")
tdSql.checkData(0,0,28)
tdSql.checkData(0,1,184)
tdSql.checkData(0,2,-99999)
tdSql.checkData(0,3,-999)
tdSql.checkData(0,4,28.000000000)
def do_distribute_spread(self):
# init
self.vnode_disbutes = None
self.ts = 1537146000000
# do
self.prepare_datas_of_distribute()
self.check_distribute_datas()
self.check_spread_status()
self.distribute_agg_query()
print("do_distribute_spread .................. [passed]")
#
# ------------------ main ------------------
#
def test_func_agg_spread(self):
""" Fun: spread()
1. Sim case including time windows, filtering on ordinary data columns, filtering on tag columns, GROUP BY, and PARTITION BY.
2. Query on super/child/normal table
3. Support types
4. Error cases
5. Query with filter conditions
6. Query with group by
7. Query with distribute aggregate
8. Check function work status
Since: v3.0.0.0
Labels: common,ci
Jira: None
History:
- 2025-4-28 Simon Guan Migrated from tsim/query/count_spread.sim
- 2025-9-28 Alex Duan Migrated from uncatalog/system-test/2-query/test_spread.py
- 2025-9-28 Alex Duan Migrated from uncatalog/system-test/2-query/test_distribute_agg_spread.py
"""
self.do_sim_spread()
self.do_spread()
self.do_distribute_spread()
Reference in a new issue View git blame Copy permalink