firebird/firebird-qa
6
0
Fork 0
mirror of https://github.com/FirebirdSQL/firebird-qa.git synced 2025-01-22 21:43:06 +01:00
Code Issues Activity
b458cecb1e
firebird-qa/tests/bugs/core_5488_test.py

214 lines
9.5 KiB
Python
Raw Blame History

#coding:utf-8
#
# id: bugs.core_5488
# title: Timeout for running SQL statement (SET STATEMENT TIMEOUT <N>)
# decription:
# ::: NB:::
#
# This is _initial_ test with trivial checks:
# 1) it hould be possible to cancel 'heavy query' executed on <this_test_DB>;
# 2) the same, but passed via ES/EDS, should also be cancelled;
# Elapsed time should be equal to statement timeout with precise -0.1% / +2%
# (minus 0.1% - because PC clock can be imprecise!)
#
# More complex cases will be implemented later.
#
# Explanations were given by hvlad, see letters from him of 05-mar-2017.
# Checked on WI-T4.0.0.546.
#
# 28.10.2019: increased spread between low and high limit for ratio elapsed_ms / sttm_timeout.
# Checked on:
# 4.0.0.1635 SS: 3.672s.
# 4.0.0.1633 CS: 3.983s.
#
# 01.04.2020: changed logic in execute block (see "--- 2 ===') after letter from Vlad, 31.03.2020 13:15.
# We must NOT suppose that engine could always in time do SUSPEND after ES/EDS will be interrupted.
# For this reason all results that we get inside EB must be stored as context variables, WITHOUT using
# output arguments which are returned by SUSPEND (see variables: 'returned_cnt', 'gds_after_eds').
# Engine restarts mechanism was changed 30-03-2020, see:
# https://github.com/FirebirdSQL/firebird/commit/059694b0b4f3a6f92f852b0a7be6358d708ee5e0
# et al.
#
# Checked 01.04.2020 on:
# 4.0.0.1850 SS: 3.686s.
# 4.0.0.1848 CS: 4.070s.
# 4.0.0.1767 SC: 3.821s.
#
# tracker_id: CORE-5488
# min_versions: ['4.0']
# versions: 4.0
# qmid: None
import pytest
from firebird.qa import db_factory, isql_act, Action
# version: 4.0
# resources: None
substitutions_1 = [('[ \t]+', ' ')]
init_script_1 = """"""
db_1 = db_factory(sql_dialect=3, init=init_script_1)
test_script_1 = """
set list on;
-- ####################################
set statement timeout 1234 millisecond;
-- ####################################
---------------------------------------- 1 ----------------------------------------------
set term ^;
execute block as
begin
rdb$set_context('USER_SESSION','EB_FINAL_DTS', null);
rdb$set_context('USER_SESSION','MIN_RATIO', 0.95); -- BEFORE 28.10.2019: 0.999
rdb$set_context('USER_SESSION','MAX_RATIO', 1.05); -- BEFORE 28.10.2019: 1.02
rdb$set_context('USER_SESSION','START_DTS', cast('now' as timestamp));
end^
set term ;^
select count(*) from rdb$types,rdb$types,rdb$types;
set term ^;
execute block as
begin
rdb$set_context('USER_SESSION','EB_FINAL_DTS', cast('now' as timestamp));
end^
set term ;^
select iif( 1.00 * elapsed_ms / sttm_timeout between low_bound and high_bound,
'Acceptable',
'WRONG: sttm_timeout = ' || sttm_timeout
|| ', elapsed_ms = ' || elapsed_ms
|| ', ratio elapsed_ms / sttm_timeout = '|| (1.00 * elapsed_ms / sttm_timeout)
||' must belong to [' || low_bound || ' ... '|| high_bound || ']'
) as result_1
from (
select
cast(rdb$get_context('SYSTEM','STATEMENT_TIMEOUT') as int) as sttm_timeout
,datediff( millisecond
from cast( rdb$get_context('USER_SESSION','START_DTS') as timestamp)
to cast(rdb$get_context('USER_SESSION','EB_FINAL_DTS') as timestamp)
) elapsed_ms
,cast( rdb$get_context('USER_SESSION','MIN_RATIO') as double precision) as low_bound -- ### THRESHOLD, LOW ###
,cast( rdb$get_context('USER_SESSION','MAX_RATIO') as double precision) as high_bound -- ### THRESHOLD, HIGH ###
from rdb$database
);
---------------------------------------- 2 ----------------------------------------------
set term ^;
execute block as
begin
rdb$set_context('USER_SESSION','EB_FINAL_DTS', null);
rdb$set_context('USER_SESSION','START_DTS', cast('now' as timestamp));
end
^
execute block as
declare p varchar(20) = 'masterkey';
declare returned_cnt int = null;
declare gds_after_eds int = null;
begin
begin
execute statement 'select count(*) from rdb$types,rdb$types,rdb$types'
on external 'localhost:' || rdb$get_context('SYSTEM','DB_NAME')
as user current_user password :p role right( uuid_to_char(gen_uuid()), 12 )
into returned_cnt;
when any do
begin
-- Control MUST come here after statement will be interruptec (cancelled) in EDS.
-- gdscode should be = 335544926.
gds_after_eds = gdscode;
rdb$set_context('USER_SESSION','GDS_IN_WHEN_ANY_BLOCK', gds_after_eds);
rdb$set_context('USER_SESSION','TIMESTAMP_IN_WHEN_ANY_BLOCK', cast('now' as timestamp));
end
end
if (gds_after_eds is null) then
begin
-- Normally we should NOT pass here, thus following context variables must remain NULL:
rdb$set_context('USER_SESSION','EB_FINAL_DTS', timestamp '01.01.2000 00:00:01' );
rdb$set_context('USER_SESSION','EB_EDS_CNT', -1);
end
-- control MUST come here, but: 'returned_cnt' will remain undefined, 'gds_after_eds' will be 335544926.
-- disabled 01.04.2020, see letter from hvlad 31.03.2020 -- suspend;
end
^
set term ;^
------------------------------------- verify results ------------------------------------
select iif( 1.00 * x.elapsed_ms / x.sttm_timeout between x.low_bound and x.high_bound,
'Acceptable',
'WRONG: sttm_timeout = ' || sttm_timeout
|| ', elapsed_ms = ' || elapsed_ms
|| ', ratio elapsed_ms / sttm_timeout = '|| (1.00 * elapsed_ms / sttm_timeout)
||' must belong to [' || low_bound || ' ... '|| high_bound || ']'
) as result_2
,x.gds_in_when_any_block -- expected: 335544926, which means that ES/EDS was really interrupted
,iif( x.end_of_eb_cnt is null, 'EXPECTED, <null>', 'UNEXPECTED: ' || x.end_of_eb_cnt ) as end_of_eb_cnt -- expected: null which means that GDSCODE in when-any block had not-null value
from (
select
cast(rdb$get_context('SYSTEM','STATEMENT_TIMEOUT') as int) as sttm_timeout
,datediff( millisecond
from cast( rdb$get_context('USER_SESSION','START_DTS') as timestamp)
to cast(rdb$get_context('USER_SESSION','TIMESTAMP_IN_WHEN_ANY_BLOCK') as timestamp)
) elapsed_ms
,rdb$get_context('USER_SESSION','GDS_IN_WHEN_ANY_BLOCK') as gds_in_when_any_block -- expected: 335544926
,rdb$get_context('USER_SESSION','EB_FINAL_DTS') end_of_eb_dts -- expected: NULL
,rdb$get_context('USER_SESSION','EB_EDS_CNT') as end_of_eb_cnt -- expected: NULL
,cast( rdb$get_context('USER_SESSION','MIN_RATIO') as double precision) as low_bound -- ### THRESHOLD, LOW ###
,cast( rdb$get_context('USER_SESSION','MAX_RATIO') as double precision) as high_bound -- ### THRESHOLD, HIGH ###
from rdb$database
) x;
commit;
-- ||||||||||||||||||||||||||||
-- ###################################||| FB 4.0+, SS and SC |||##############################
-- ||||||||||||||||||||||||||||
-- If we check SS or SC and ExtConnPoolLifeTime > 0 (config parameter FB 4.0+) then current
-- DB (bugs.core_NNNN.fdb) will be 'captured' by firebird.exe process and fbt_run utility
-- will not able to drop this database at the final point of test.
-- Moreover, DB file will be hold until all activity in firebird.exe completed and AFTER this
-- we have to wait for <ExtConnPoolLifeTime> seconds after it (discussion and small test see
-- in the letter to hvlad and dimitr 13.10.2019 11:10).
-- This means that one need to kill all connections to prevent from exception on cleanup phase:
-- SQLCODE: -901 / lock time-out on wait transaction / object <this_test_DB> is in use
-- #############################################################################################
delete from mon$attachments where mon$attachment_id != current_connection;
commit;
"""
act_1 = isql_act('db_1', test_script_1, substitutions=substitutions_1)
expected_stdout_1 = """
RESULT_1 Acceptable
RESULT_2 Acceptable
GDS_IN_WHEN_ANY_BLOCK 335544926
END_OF_EB_CNT EXPECTED, <null>
"""
expected_stderr_1 = """
Statement failed, SQLSTATE = HY008
operation was cancelled
-Attachment level timeout expired.
"""
@pytest.mark.version('>=4.0')
def test_1(act_1: Action):
act_1.expected_stdout = expected_stdout_1
act_1.expected_stderr = expected_stderr_1
act_1.execute()
assert act_1.clean_expected_stderr == act_1.clean_stderr
assert act_1.clean_expected_stdout == act_1.clean_stdout
Reference in New Issue View Git Blame Copy Permalink