2 @message |
assert
checked_mode: table, STDLOG: Records affected: 4
checked_mode: table, STDLOG: ID
checked_mode: table, STDLOG:
checked_mode: table, STDLOG: -5
checked_mode: table, STDLOG: -2
checked_mode: table, STDLOG: -1
checked_mode: table, STDLOG: 3
checked_mode: table, STDLOG: 4
checked_mode: table, STDLOG: Records affected: 5
checked_mode: table, STDLOG: OLD_ID OP SNAP_NO_RANK
checked_mode: table, STDLOG:
- checked_mode: table, STDLOG: 2 UPD 1
+ checked_mode: table, STDLOG: 2UPD 1
- checked_mode: table, STDLOG: 2 UPD 2
+ checked_mode: table, STDLOG: 2UPD 2
- checked_mode: table, STDLOG: 1 UPD 2
+ checked_mode: table, STDLOG: 1UPD 2
- checked_mode: table, STDLOG: 2 UPD 3
+ checked_mode: table, STDLOG: 2UPD 3
- checked_mode: table, STDLOG: 1 UPD 3
+ checked_mode: table, STDLOG: 1UPD 3
- checked_mode: table, STDLOG: 2 UPD 4
+ checked_mode: table, STDLOG: 2UPD 4
- checked_mode: table, STDLOG: 1 UPD 4
+ checked_mode: table, STDLOG: 1UPD 4
- checked_mode: table, STDLOG: -3 UPD 4
+ checked_mode: table, STDLOG: -3UPD 4
- checked_mode: table, STDLOG: -4 UPD 4
+ checked_mode: table, STDLOG: -4UPD 4
checked_mode: table, STDLOG: Records affected: 9
LOG DETAILS:
2025-07-01 06:05:13.598
2025-07-01 06:05:13.606 act = <firebird.qa.plugin.Action object at [hex]>
2025-07-01 06:05:13.618 fn_worker_sql = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.sql')
2025-07-01 06:05:13.629 fn_worker_log = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.log')
2025-07-01 06:05:13.641 fn_worker_err = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.err')
2025-07-01 06:05:13.652 capsys = <_pytest.capture.CaptureFixture object at [hex]>
2025-07-01 06:05:13.660
2025-07-01 06:05:13.667 @pytest.mark.trace
2025-07-01 06:05:13.675 @pytest.mark.version('>=4.0')
2025-07-01 06:05:13.683 def test_1(act: Action, fn_worker_sql: Path, fn_worker_log: Path, fn_worker_err: Path, capsys):
2025-07-01 06:05:13.690 sql_init = (act.files_dir / 'read-consist-sttm-restart-DDL.sql').read_text()
2025-07-01 06:05:13.697
2025-07-01 06:05:13.703 for checked_mode in('table', 'view'):
2025-07-01 06:05:13.709 target_obj = 'test' if checked_mode == 'table' else 'v_test'
2025-07-01 06:05:13.715
2025-07-01 06:05:13.721 SQL_TO_BE_RESTARTED = f"update /* {SQL_TAG_THAT_WE_WAITING_FOR} */ {target_obj} set id = -id where id <= 2 order by id DESC rows 4"
2025-07-01 06:05:13.726
2025-07-01 06:05:13.732 sql_addi = f'''
2025-07-01 06:05:13.737 set term ^;
2025-07-01 06:05:13.743 execute block as
2025-07-01 06:05:13.750 begin
2025-07-01 06:05:13.756 rdb$set_context('USER_SESSION', 'WHO', 'INIT_DATA');
2025-07-01 06:05:13.761 end
2025-07-01 06:05:13.767 ^
2025-07-01 06:05:13.773 set term ;^
2025-07-01 06:05:13.779 insert into {target_obj}(id, x)
2025-07-01 06:05:13.786 select row_number()over(),row_number()over()
2025-07-01 06:05:13.794 from rdb$types rows 5;
2025-07-01 06:05:13.803 commit;
2025-07-01 06:05:13.810 '''
2025-07-01 06:05:13.816
2025-07-01 06:05:13.822 act.isql(switches=['-q'], input = ''.join( (sql_init, sql_addi) ))
2025-07-01 06:05:13.830 # ::: NOTE ::: We have to immediately quit if any error raised in prepare phase.
2025-07-01 06:05:13.839 # See also letter from dimitr, 01-feb-2022 14:46
2025-07-01 06:05:13.845 assert act.stderr == ''
2025-07-01 06:05:13.853 act.reset()
2025-07-01 06:05:13.859
2025-07-01 06:05:13.865 trace_cfg_items = [
2025-07-01 06:05:13.871 'time_threshold = 0',
2025-07-01 06:05:13.879 'log_errors = true',
2025-07-01 06:05:13.890 'log_statement_start = true',
2025-07-01 06:05:13.898 'log_statement_finish = true',
2025-07-01 06:05:13.909 ]
2025-07-01 06:05:13.920
2025-07-01 06:05:13.930 with act.trace(db_events = trace_cfg_items, encoding=locale.getpreferredencoding()):
2025-07-01 06:05:13.939
2025-07-01 06:05:13.947 with act.db.connect() as con_lock_1, act.db.connect() as con_lock_2, act.db.connect() as con_monitoring:
2025-07-01 06:05:13.955
2025-07-01 06:05:13.961 tpb_monitoring = tpb(isolation=Isolation.READ_COMMITTED_RECORD_VERSION, lock_timeout=0)
2025-07-01 06:05:13.967 tx_monitoring = con_monitoring.transaction_manager(tpb_monitoring)
2025-07-01 06:05:13.973 cur_monitoring = tx_monitoring.cursor()
2025-07-01 06:05:13.979
2025-07-01 06:05:13.987 for i,c in enumerate((con_lock_1,con_lock_2)):
2025-07-01 06:05:13.994 sttm = f"execute block as begin rdb$set_context('USER_SESSION', 'WHO', 'LOCKER #{i+1}'); end"
2025-07-01 06:05:14.004 c.execute_immediate(sttm)
2025-07-01 06:05:14.014
2025-07-01 06:05:14.022
2025-07-01 06:05:14.029 #########################
2025-07-01 06:05:14.037 ### L O C K E R - 1 ###
2025-07-01 06:05:14.044 #########################
2025-07-01 06:05:14.050
2025-07-01 06:05:14.056 con_lock_1.execute_immediate( f'update {target_obj} set id=id where id=1' )
2025-07-01 06:05:14.068
2025-07-01 06:05:14.077 worker_sql = f'''
2025-07-01 06:05:14.085 set list on;
2025-07-01 06:05:14.092 set autoddl off;
2025-07-01 06:05:14.099 set term ^;
2025-07-01 06:05:14.104 execute block returns (whoami varchar(30)) as
2025-07-01 06:05:14.110 begin
2025-07-01 06:05:14.117 whoami = 'WORKER'; -- , ATT#' || current_connection;
2025-07-01 06:05:14.123 rdb$set_context('USER_SESSION','WHO', whoami);
2025-07-01 06:05:14.130 -- suspend;
2025-07-01 06:05:14.136 end
2025-07-01 06:05:14.142 ^
2025-07-01 06:05:14.148 set term ;^
2025-07-01 06:05:14.154 commit;
2025-07-01 06:05:14.161 SET KEEP_TRAN_PARAMS ON;
2025-07-01 06:05:14.167 set transaction read committed read consistency;
2025-07-01 06:05:14.173 --select current_connection, current_transaction from rdb$database;
2025-07-01 06:05:14.179 set list off;
2025-07-01 06:05:14.186 set wng off;
2025-07-01 06:05:14.192
2025-07-01 06:05:14.198 --set plan on;
2025-07-01 06:05:14.207 set count on;
2025-07-01 06:05:14.213
2025-07-01 06:05:14.219 -- this must hang because of locker-1:
2025-07-01 06:05:14.225 {SQL_TO_BE_RESTARTED};
2025-07-01 06:05:14.230
2025-07-01 06:05:14.236 -- check results:
2025-07-01 06:05:14.242 -- ###############
2025-07-01 06:05:14.248 select id from {target_obj} order by id; -- one record must remain, with ID = -5
2025-07-01 06:05:14.254
2025-07-01 06:05:14.261 select v.old_id, v.op, v.snap_no_rank -- snap_no_rank must have four unique values: 1,2,3 and 4.
2025-07-01 06:05:14.272 from v_worker_log v
2025-07-01 06:05:14.278 where v.op = 'upd';
2025-07-01 06:05:14.284
2025-07-01 06:05:14.290 --set width who 10;
2025-07-01 06:05:14.297 -- DO NOT check this! Values can differ here from one run to another!
2025-07-01 06:05:14.303 -- select id, trn, who, old_id, new_id, op, rec_vers, global_cn, snap_no from tlog_done order by id;
2025-07-01 06:05:14.309 rollback;
2025-07-01 06:05:14.316
2025-07-01 06:05:14.325 '''
2025-07-01 06:05:14.338
2025-07-01 06:05:14.350 fn_worker_sql.write_text(worker_sql)
2025-07-01 06:05:14.363
2025-07-01 06:05:14.373 with fn_worker_log.open(mode='w') as hang_out, fn_worker_err.open(mode='w') as hang_err:
2025-07-01 06:05:14.381
2025-07-01 06:05:14.395 ############################################################################
2025-07-01 06:05:14.405 ### L A U N C H W O R K E R U S I N G I S Q L, A S Y N C. ###
2025-07-01 06:05:14.414 ############################################################################
2025-07-01 06:05:14.420 p_worker = subprocess.Popen([act.vars['isql'], '-i', str(fn_worker_sql),
2025-07-01 06:05:14.427 '-user', act.db.user,
2025-07-01 06:05:14.435 '-password', act.db.password,
2025-07-01 06:05:14.442 '-pag', '999999',
2025-07-01 06:05:14.451 act.db.dsn
2025-07-01 06:05:14.463 ],
2025-07-01 06:05:14.473 stdout = hang_out,
2025-07-01 06:05:14.480 stderr = hang_err
2025-07-01 06:05:14.486 )
2025-07-01 06:05:14.491
2025-07-01 06:05:14.497 # NB: when ISQL will establish attach, first record that it must lock is ID = 2 -- see above SQL_TO_BE_RESTARTED
2025-07-01 06:05:14.502 # We must to ensure that this (worker) attachment has been really created and LOCKS this record:
2025-07-01 06:05:14.508 #
2025-07-01 06:05:14.515 wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id=2', SQL_TAG_THAT_WE_WAITING_FOR)
2025-07-01 06:05:14.521
2025-07-01 06:05:14.529 #########################
2025-07-01 06:05:14.537 ### L O C K E R - 2 ###
2025-07-01 06:05:14.543 #########################
2025-07-01 06:05:14.550
2025-07-01 06:05:14.558 # Change ID so that it **will* be included in the set of rows that must be affected by session-worker:
2025-07-01 06:05:14.565 con_lock_2.execute_immediate( f'update {target_obj} set id = -5 where abs(id) = 5;' )
2025-07-01 06:05:14.572 con_lock_2.commit()
2025-07-01 06:05:14.579 con_lock_2.execute_immediate( f'update {target_obj} set id = id where abs(id) = 5;' )
2025-07-01 06:05:14.586
2025-07-01 06:05:14.595 #########################
2025-07-01 06:05:14.606 ### L O C K E R - 1 ###
2025-07-01 06:05:14.614 #########################
2025-07-01 06:05:14.621
2025-07-01 06:05:14.627 con_lock_1.commit() # releases record with ID=1 ==> now it can be locked by worker.
2025-07-01 06:05:14.631
2025-07-01 06:05:14.636 # We have to WAIT HERE until worker will actually 'catch' just released record with ID = 1.
2025-07-01 06:05:14.641 #
2025-07-01 06:05:14.646 wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id=1', SQL_TAG_THAT_WE_WAITING_FOR)
2025-07-01 06:05:14.651
2025-07-01 06:05:14.656 # If we come here then it means that record with ID = 1 for sure is locked by WORKER.
2025-07-01 06:05:14.662
2025-07-01 06:05:14.668 # Change ID so that it **will* be included in the set of rows that must be affected by session-worker:
2025-07-01 06:05:14.673 con_lock_1.execute_immediate( f'update {target_obj} set id = -4 where abs(id) = 4;' )
2025-07-01 06:05:14.679 con_lock_1.commit()
2025-07-01 06:05:14.685 con_lock_1.execute_immediate( f'update {target_obj} set id = id where abs(id) = 4;' )
2025-07-01 06:05:14.691
2025-07-01 06:05:14.699
2025-07-01 06:05:14.710 #########################
2025-07-01 06:05:14.719 ### L O C K E R - 2 ###
2025-07-01 06:05:14.726 #########################
2025-07-01 06:05:14.731
2025-07-01 06:05:14.737 con_lock_2.commit() # releases record with ID = -5, but session-worker is waiting for record with ID = -4 (that was changed by locker-1).
2025-07-01 06:05:14.742
2025-07-01 06:05:14.747 # We have to WAIT HERE until worker will actually 'catch' just released record with ID = -5:
2025-07-01 06:05:14.753 #
2025-07-01 06:05:14.759 wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id = -5', SQL_TAG_THAT_WE_WAITING_FOR)
2025-07-01 06:05:14.764
2025-07-01 06:05:14.770 # If we come here then it means that record with ID = -5 for sure is locked by WORKER.
2025-07-01 06:05:14.776
2025-07-01 06:05:14.783 con_lock_2.execute_immediate( f'update {target_obj} set id = -3 where abs(id) = 3;' )
2025-07-01 06:05:14.790 con_lock_2.commit()
2025-07-01 06:05:14.797 con_lock_2.execute_immediate( f'update {target_obj} set id = id where abs(id) = 3;' )
2025-07-01 06:05:14.804
2025-07-01 06:05:14.811 #########################
2025-07-01 06:05:14.818 ### L O C K E R - 1 ###
2025-07-01 06:05:14.825 #########################
2025-07-01 06:05:14.832
2025-07-01 06:05:14.844 con_lock_1.commit() # This releases row with ID=-4 but session-worker is waiting for ID = - 3 (changed by locker-2).
2025-07-01 06:05:14.854
2025-07-01 06:05:14.861 # We have to WAIT HERE until worker will actually 'catch' just released record with ID = -4:
2025-07-01 06:05:14.868 #
2025-07-01 06:05:14.876 wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id = -4', SQL_TAG_THAT_WE_WAITING_FOR)
2025-07-01 06:05:14.882
2025-07-01 06:05:14.888 # If we come here then it means that record with ID = -4 for sure is locked by WORKER.
2025-07-01 06:05:14.895
2025-07-01 06:05:14.902 con_lock_2.commit() # This releases row with ID=-3. No more locked rows so session-worker can finish its mission.
2025-07-01 06:05:14.908
2025-07-01 06:05:14.913 # Here we wait for ISQL complete its mission:
2025-07-01 06:05:14.919 p_worker.wait()
2025-07-01 06:05:14.925
2025-07-01 06:05:14.935 #< with act.db.connect()
2025-07-01 06:05:14.941
2025-07-01 06:05:14.949 for g in (fn_worker_log, fn_worker_err):
2025-07-01 06:05:14.955 with g.open() as f:
2025-07-01 06:05:14.962 for line in f:
2025-07-01 06:05:14.969 if line.split():
2025-07-01 06:05:14.976 if g == fn_worker_log:
2025-07-01 06:05:14.983 print(f'checked_mode: {checked_mode}, STDLOG: {line}')
2025-07-01 06:05:14.990 else:
2025-07-01 06:05:14.997 print(f'UNEXPECTED STDERR {line}')
2025-07-01 06:05:15.003
2025-07-01 06:05:15.011 expected_stdout_worker = f"""
2025-07-01 06:05:15.018 checked_mode: {checked_mode}, STDLOG: Records affected: 4
2025-07-01 06:05:15.023
2025-07-01 06:05:15.030 checked_mode: {checked_mode}, STDLOG: ID
2025-07-01 06:05:15.036 checked_mode: {checked_mode}, STDLOG: =======
2025-07-01 06:05:15.041 checked_mode: {checked_mode}, STDLOG: -5
2025-07-01 06:05:15.047 checked_mode: {checked_mode}, STDLOG: -2
2025-07-01 06:05:15.053 checked_mode: {checked_mode}, STDLOG: -1
2025-07-01 06:05:15.059 checked_mode: {checked_mode}, STDLOG: 3
2025-07-01 06:05:15.065 checked_mode: {checked_mode}, STDLOG: 4
2025-07-01 06:05:15.071 checked_mode: {checked_mode}, STDLOG: Records affected: 5
2025-07-01 06:05:15.076
2025-07-01 06:05:15.082 checked_mode: {checked_mode}, STDLOG: OLD_ID OP SNAP_NO_RANK
2025-07-01 06:05:15.088 checked_mode: {checked_mode}, STDLOG: ======= ====== =====================
2025-07-01 06:05:15.100 checked_mode: {checked_mode}, STDLOG: 2 UPD 1
2025-07-01 06:05:15.107 checked_mode: {checked_mode}, STDLOG: 2 UPD 2
2025-07-01 06:05:15.112 checked_mode: {checked_mode}, STDLOG: 1 UPD 2
2025-07-01 06:05:15.117 checked_mode: {checked_mode}, STDLOG: 2 UPD 3
2025-07-01 06:05:15.123 checked_mode: {checked_mode}, STDLOG: 1 UPD 3
2025-07-01 06:05:15.129 checked_mode: {checked_mode}, STDLOG: 2 UPD 4
2025-07-01 06:05:15.135 checked_mode: {checked_mode}, STDLOG: 1 UPD 4
2025-07-01 06:05:15.141 checked_mode: {checked_mode}, STDLOG: -3 UPD 4
2025-07-01 06:05:15.147 checked_mode: {checked_mode}, STDLOG: -4 UPD 4
2025-07-01 06:05:15.153 checked_mode: {checked_mode}, STDLOG: Records affected: 9
2025-07-01 06:05:15.159 """
2025-07-01 06:05:15.165 act.expected_stdout = expected_stdout_worker
2025-07-01 06:05:15.170 act.stdout = capsys.readouterr().out
2025-07-01 06:05:15.176 > assert act.clean_stdout == act.clean_expected_stdout
2025-07-01 06:05:15.182 E assert
2025-07-01 06:05:15.189 E checked_mode: table, STDLOG: Records affected: 4
2025-07-01 06:05:15.195 E checked_mode: table, STDLOG: ID
2025-07-01 06:05:15.201 E checked_mode: table, STDLOG:
2025-07-01 06:05:15.207 E checked_mode: table, STDLOG: -5
2025-07-01 06:05:15.213 E checked_mode: table, STDLOG: -2
2025-07-01 06:05:15.219 E checked_mode: table, STDLOG: -1
2025-07-01 06:05:15.225 E checked_mode: table, STDLOG: 3
2025-07-01 06:05:15.230 E checked_mode: table, STDLOG: 4
2025-07-01 06:05:15.236 E checked_mode: table, STDLOG: Records affected: 5
2025-07-01 06:05:15.241 E checked_mode: table, STDLOG: OLD_ID OP SNAP_NO_RANK
2025-07-01 06:05:15.247 E checked_mode: table, STDLOG:
2025-07-01 06:05:15.254 E - checked_mode: table, STDLOG: 2 UPD 1
2025-07-01 06:05:15.266 E + checked_mode: table, STDLOG: 2UPD 1
2025-07-01 06:05:15.271 E - checked_mode: table, STDLOG: 2 UPD 2
2025-07-01 06:05:15.283 E + checked_mode: table, STDLOG: 2UPD 2
2025-07-01 06:05:15.289 E - checked_mode: table, STDLOG: 1 UPD 2
2025-07-01 06:05:15.306 E + checked_mode: table, STDLOG: 1UPD 2
2025-07-01 06:05:15.315 E - checked_mode: table, STDLOG: 2 UPD 3
2025-07-01 06:05:15.327 E + checked_mode: table, STDLOG: 2UPD 3
2025-07-01 06:05:15.332 E - checked_mode: table, STDLOG: 1 UPD 3
2025-07-01 06:05:15.341 E + checked_mode: table, STDLOG: 1UPD 3
2025-07-01 06:05:15.346 E - checked_mode: table, STDLOG: 2 UPD 4
2025-07-01 06:05:15.355 E + checked_mode: table, STDLOG: 2UPD 4
2025-07-01 06:05:15.360 E - checked_mode: table, STDLOG: 1 UPD 4
2025-07-01 06:05:15.369 E + checked_mode: table, STDLOG: 1UPD 4
2025-07-01 06:05:15.373 E - checked_mode: table, STDLOG: -3 UPD 4
2025-07-01 06:05:15.383 E + checked_mode: table, STDLOG: -3UPD 4
2025-07-01 06:05:15.387 E - checked_mode: table, STDLOG: -4 UPD 4
2025-07-01 06:05:15.397 E + checked_mode: table, STDLOG: -4UPD 4
2025-07-01 06:05:15.401 E checked_mode: table, STDLOG: Records affected: 9
2025-07-01 06:05:15.406
2025-07-01 06:05:15.411 tests/functional/transactions/test_read_consist_sttm_restart_on_update_04.py:442: AssertionError
2025-07-01 06:05:15.415 ---------------------------- Captured stdout setup -----------------------------
2025-07-01 06:05:15.420 Creating db: localhost:/var/tmp/qa_2024/test_12487/test.fdb [page_size=None, sql_dialect=None, charset='NONE', user=SYSDBA, password=masterkey]
|
3 #text |
act = <firebird.qa.plugin.Action pytest object at [hex]>
fn_worker_sql = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.sql')
fn_worker_log = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.log')
fn_worker_err = PosixPath('/var/tmp/qa_2024/test_12487/tmp_worker.err')
capsys = <_pytest.capture.CaptureFixture pytest object at [hex]>
@pytest.mark.trace
@pytest.mark.version('>=4.0')
def test_1(act: Action, fn_worker_sql: Path, fn_worker_log: Path, fn_worker_err: Path, capsys):
sql_init = (act.files_dir / 'read-consist-sttm-restart-DDL.sql').read_text()
for checked_mode in('table', 'view'):
target_obj = 'test' if checked_mode == 'table' else 'v_test'
SQL_TO_BE_RESTARTED = f"update /* {SQL_TAG_THAT_WE_WAITING_FOR} */ {target_obj} set id = -id where id <= 2 order by id DESC rows 4"
sql_addi = f'''
set term ^;
execute block as
begin
rdb$set_context('USER_SESSION', 'WHO', 'INIT_DATA');
end
^
set term ;^
insert into {target_obj}(id, x)
select row_number()over(),row_number()over()
from rdb$types rows 5;
commit;
'''
act.isql(switches=['-q'], input = ''.join( (sql_init, sql_addi) ))
# ::: NOTE ::: We have to immediately quit if any error raised in prepare phase.
# See also letter from dimitr, 01-feb-2022 14:46
assert act.stderr == ''
act.reset()
trace_cfg_items = [
'time_threshold = 0',
'log_errors = true',
'log_statement_start = true',
'log_statement_finish = true',
]
with act.trace(db_events = trace_cfg_items, encoding=locale.getpreferredencoding()):
with act.db.connect() as con_lock_1, act.db.connect() as con_lock_2, act.db.connect() as con_monitoring:
tpb_monitoring = tpb(isolation=Isolation.READ_COMMITTED_RECORD_VERSION, lock_timeout=0)
tx_monitoring = con_monitoring.transaction_manager(tpb_monitoring)
cur_monitoring = tx_monitoring.cursor()
for i,c in enumerate((con_lock_1,con_lock_2)):
sttm = f"execute block as begin rdb$set_context('USER_SESSION', 'WHO', 'LOCKER #{i+1}'); end"
c.execute_immediate(sttm)
#########################
### L O C K E R - 1 ###
#########################
con_lock_1.execute_immediate( f'update {target_obj} set id=id where id=1' )
worker_sql = f'''
set list on;
set autoddl off;
set term ^;
execute block returns (whoami varchar(30)) as
begin
whoami = 'WORKER'; -- , ATT#' || current_connection;
rdb$set_context('USER_SESSION','WHO', whoami);
-- suspend;
end
^
set term ;^
commit;
SET KEEP_TRAN_PARAMS ON;
set transaction read committed read consistency;
--select current_connection, current_transaction from rdb$database;
set list off;
set wng off;
--set plan on;
set count on;
-- this must hang because of locker-1:
{SQL_TO_BE_RESTARTED};
-- check results:
-- ###############
select id from {target_obj} order by id; -- one record must remain, with ID = -5
select v.old_id, v.op, v.snap_no_rank -- snap_no_rank must have four unique values: 1,2,3 and 4.
from v_worker_log v
where v.op = 'upd';
--set width who 10;
-- DO NOT check this! Values can differ here from one run to another!
-- select id, trn, who, old_id, new_id, op, rec_vers, global_cn, snap_no from tlog_done order by id;
rollback;
'''
fn_worker_sql.write_text(worker_sql)
with fn_worker_log.open(mode='w') as hang_out, fn_worker_err.open(mode='w') as hang_err:
############################################################################
### L A U N C H W O R K E R U S I N G I S Q L, A S Y N C. ###
############################################################################
p_worker = subprocess.Popen([act.vars['isql'], '-i', str(fn_worker_sql),
'-user', act.db.user,
'-password', act.db.password,
'-pag', '999999',
act.db.dsn
],
stdout = hang_out,
stderr = hang_err
)
# NB: when ISQL will establish attach, first record that it must lock is ID = 2 -- see above SQL_TO_BE_RESTARTED
# We must to ensure that this (worker) attachment has been really created and LOCKS this record:
#
wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id=2', SQL_TAG_THAT_WE_WAITING_FOR)
#########################
### L O C K E R - 2 ###
#########################
# Change ID so that it **will* be included in the set of rows that must be affected by session-worker:
con_lock_2.execute_immediate( f'update {target_obj} set id = -5 where abs(id) = 5;' )
con_lock_2.commit()
con_lock_2.execute_immediate( f'update {target_obj} set id = id where abs(id) = 5;' )
#########################
### L O C K E R - 1 ###
#########################
con_lock_1.commit() # releases record with ID=1 ==> now it can be locked by worker.
# We have to WAIT HERE until worker will actually 'catch' just released record with ID = 1.
#
wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id=1', SQL_TAG_THAT_WE_WAITING_FOR)
# If we come here then it means that record with ID = 1 for sure is locked by WORKER.
# Change ID so that it **will* be included in the set of rows that must be affected by session-worker:
con_lock_1.execute_immediate( f'update {target_obj} set id = -4 where abs(id) = 4;' )
con_lock_1.commit()
con_lock_1.execute_immediate( f'update {target_obj} set id = id where abs(id) = 4;' )
#########################
### L O C K E R - 2 ###
#########################
con_lock_2.commit() # releases record with ID = -5, but session-worker is waiting for record with ID = -4 (that was changed by locker-1).
# We have to WAIT HERE until worker will actually 'catch' just released record with ID = -5:
#
wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id = -5', SQL_TAG_THAT_WE_WAITING_FOR)
# If we come here then it means that record with ID = -5 for sure is locked by WORKER.
con_lock_2.execute_immediate( f'update {target_obj} set id = -3 where abs(id) = 3;' )
con_lock_2.commit()
con_lock_2.execute_immediate( f'update {target_obj} set id = id where abs(id) = 3;' )
#########################
### L O C K E R - 1 ###
#########################
con_lock_1.commit() # This releases row with ID=-4 but session-worker is waiting for ID = - 3 (changed by locker-2).
# We have to WAIT HERE until worker will actually 'catch' just released record with ID = -4:
#
wait_for_record_become_locked(tx_monitoring, cur_monitoring, f'update {target_obj} set id=id where id = -4', SQL_TAG_THAT_WE_WAITING_FOR)
# If we come here then it means that record with ID = -4 for sure is locked by WORKER.
con_lock_2.commit() # This releases row with ID=-3. No more locked rows so session-worker can finish its mission.
# Here we wait for ISQL complete its mission:
p_worker.wait()
#< with act.db.connect()
for g in (fn_worker_log, fn_worker_err):
with g.open() as f:
for line in f:
if line.split():
if g == fn_worker_log:
print(f'checked_mode: {checked_mode}, STDLOG: {line}')
else:
print(f'UNEXPECTED STDERR {line}')
expected_stdout_worker = f"""
checked_mode: {checked_mode}, STDLOG: Records affected: 4
checked_mode: {checked_mode}, STDLOG: ID
checked_mode: {checked_mode}, STDLOG: =======
checked_mode: {checked_mode}, STDLOG: -5
checked_mode: {checked_mode}, STDLOG: -2
checked_mode: {checked_mode}, STDLOG: -1
checked_mode: {checked_mode}, STDLOG: 3
checked_mode: {checked_mode}, STDLOG: 4
checked_mode: {checked_mode}, STDLOG: Records affected: 5
checked_mode: {checked_mode}, STDLOG: OLD_ID OP SNAP_NO_RANK
checked_mode: {checked_mode}, STDLOG: ======= ====== =====================
checked_mode: {checked_mode}, STDLOG: 2 UPD 1
checked_mode: {checked_mode}, STDLOG: 2 UPD 2
checked_mode: {checked_mode}, STDLOG: 1 UPD 2
checked_mode: {checked_mode}, STDLOG: 2 UPD 3
checked_mode: {checked_mode}, STDLOG: 1 UPD 3
checked_mode: {checked_mode}, STDLOG: 2 UPD 4
checked_mode: {checked_mode}, STDLOG: 1 UPD 4
checked_mode: {checked_mode}, STDLOG: -3 UPD 4
checked_mode: {checked_mode}, STDLOG: -4 UPD 4
checked_mode: {checked_mode}, STDLOG: Records affected: 9
"""
act.expected_stdout = expected_stdout_worker
act.stdout = capsys.readouterr().out
> assert act.clean_stdout == act.clean_expected_stdout
E assert
E checked_mode: table, STDLOG: Records affected: 4
E checked_mode: table, STDLOG: ID
E checked_mode: table, STDLOG:
E checked_mode: table, STDLOG: -5
E checked_mode: table, STDLOG: -2
E checked_mode: table, STDLOG: -1
E checked_mode: table, STDLOG: 3
E checked_mode: table, STDLOG: 4
E checked_mode: table, STDLOG: Records affected: 5
E checked_mode: table, STDLOG: OLD_ID OP SNAP_NO_RANK
E checked_mode: table, STDLOG:
E - checked_mode: table, STDLOG: 2 UPD 1
E + checked_mode: table, STDLOG: 2UPD 1
E - checked_mode: table, STDLOG: 2 UPD 2
E + checked_mode: table, STDLOG: 2UPD 2
E - checked_mode: table, STDLOG: 1 UPD 2
E + checked_mode: table, STDLOG: 1UPD 2
E - checked_mode: table, STDLOG: 2 UPD 3
E + checked_mode: table, STDLOG: 2UPD 3
E - checked_mode: table, STDLOG: 1 UPD 3
E + checked_mode: table, STDLOG: 1UPD 3
E - checked_mode: table, STDLOG: 2 UPD 4
E + checked_mode: table, STDLOG: 2UPD 4
E - checked_mode: table, STDLOG: 1 UPD 4
E + checked_mode: table, STDLOG: 1UPD 4
E - checked_mode: table, STDLOG: -3 UPD 4
E + checked_mode: table, STDLOG: -3UPD 4
E - checked_mode: table, STDLOG: -4 UPD 4
E + checked_mode: table, STDLOG: -4UPD 4
E checked_mode: table, STDLOG: Records affected: 9
tests/functional/transactions/test_read_consist_sttm_restart_on_update_04.py:442: AssertionError
|