zettadb / kunlun-storage-081809 Goto Github PK

Issue migrated from trac ticket # 145

component: kunlun-storage | priority: major | resolution: fixed

2021-08-25 10:29:07: @jd-zhang created the issue

Currently, we use pgx as user name for storage nodes to accept connections from computing nodes remotely, and also use this account to accept other admin connections.

Sometimes we need to admin the binary logs, but this account does not have the priority to do this.

We need to add BINLOG_ADMIN priority to this user.

Issue migrated from trac ticket # 11

component: kunlun-storage | priority: major

2021-03-15 11:35:42: @jd-zhang created the issue

The cluster:
One meta shard and two data shards.

• Create 15 tables

• first shard has 9 tables
• second shard has 6 tables
• create 255 tables, and all these tables fall into the same data shard.

The command to create 255 table is:
i=0
while test $i -lt 255; do psql -h localhost -p 5401 -Ukunlun postgres -c "create table batch$i(id integer primary key, f1 int)"; let i++; done

Issue migrated from trac ticket # 350

component: kunlun-storage | priority: major

2021-12-22 16:25:12: @snowao created the issue

Additional fields added like extra_info in mysql.general_log table which is attributed to the test case failed.

Issue migrated from trac ticket # 104

component: kunlun-storage | priority: major

2021-06-16 10:54:35: @jd-zhang created the issue

sql code:

DROP TABLE if exists TIMETZ_TBL;
CREATE TABLE TIMETZ_TBL (f1 time(2) with time zone);
INSERT INTO TIMETZ_TBL VALUES ('00:01 PDT');
INSERT INTO TIMETZ_TBL VALUES ('01:00 PDT');
INSERT INTO TIMETZ_TBL VALUES ('02:03 PDT');
INSERT INTO TIMETZ_TBL VALUES ('07:07 PST');
INSERT INTO TIMETZ_TBL VALUES ('08:08 EDT');
INSERT INTO TIMETZ_TBL VALUES ('11:59 PDT');
INSERT INTO TIMETZ_TBL VALUES ('12:00 PDT');
INSERT INTO TIMETZ_TBL VALUES ('12:01 PDT');
INSERT INTO TIMETZ_TBL VALUES ('23:59 PDT');
INSERT INTO TIMETZ_TBL VALUES ('11:59:59.99 PM PDT');
INSERT INTO TIMETZ_TBL VALUES ('2003-03-07 15:36:39 America/New_York');
INSERT INTO TIMETZ_TBL VALUES ('2003-07-07 15:36:39 America/New_York');

select * from TIMETZ_TBL;
tdb1=# select * from TIMETZ_TBL;
f1
00:01:00+00
01:00:00+00
02:03:00+00
07:07:00+00
08:08:00+00
11:59:00+00
12:00:00+00
12:01:00+00
23:59:00+00
23:59:59.99+00
15:36:39+00
15:36:39+00

All the timezone info is removed when putting into the table.

Issue migrated from trac ticket # 241

component: kunlun-storage | priority: minor | resolution: cant-reproduce

2021-11-23 10:00:36: @jd-zhang created the issue

When using mysqldmin to shutdown the instance, the process crashes then quits, instead of normal quits.

stack:

(gdb) bt
#0 0x00007fc0b80f88d5 in pthread_kill () from /lib64/libpthread.so.0
#1 0x00000000044c76b2 in my_write_core (sig=6) at /home/kunlun/debugbuild/percona-server-8.0.18-9/mysys/stacktrace.cc:305
#2 0x000000000343136d in handle_fatal_signal (sig=6) at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/signal_handler.cc:172
#3
#4 0x00007fc0b61bf37f in raise () from /lib64/libc.so.6
#5 0x00007fc0b61a9db5 in abort () from /lib64/libc.so.6
#6 0x00000000048fd16e in ut_dbg_assertion_failed (
expr=0x646ecc8 "buf_flush_active_lru_managers() ## srv_buf_pool_instances || buf_flush_active_lru_managers() 0",
file=0x646dee8 "/home/kunlun/debugbuild/percona-server-8.0.18-9/storage/innobase/srv/srv0start.cc", line=3531)
at /home/kunlun/debugbuild/percona-server-8.0.18-9/storage/innobase/ut/ut0dbg.cc:98
#7 0x0000000004874ce5 in srv_shutdown_page_cleaners ()
at /home/kunlun/debugbuild/percona-server-8.0.18-9/storage/innobase/srv/srv0start.cc:3530
#8 0x00000000048756c9 in srv_shutdown () at /home/kunlun/debugbuild/percona-server-8.0.18-9/storage/innobase/srv/srv0start.cc:3759
#9 0x00000000045b2981 in innodb_shutdown ()
at /home/kunlun/debugbuild/percona-server-8.0.18-9/storage/innobase/handler/ha_innodb.cc:1621
#10 0x0000000002e20e6b in ha_finalize_handlerton (plugin=0x7fc0a3feabb8)
at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/handler.cc:746
#11 0x0000000003267dd7 in plugin_deinitialize (plugin=0x7fc0a3feabb8, ref_check=true)
at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/sql_plugin.cc:1101
#12 0x00000000032684a1 in reap_plugins () at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/sql_plugin.cc:1176
#13 0x000000000326b1ed in plugin_shutdown () at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/sql_plugin.cc:2017
#14 0x00000000030148c4 in clean_up (print_message=true) at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/mysqld.cc:2292
#15 0x0000000003025b18 in mysqld_main (argc=185, argv=0x7fc0b58c31b8)
at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/mysqld.cc:7514
#16 0x0000000002dffa17 in main (argc=10, argv=0x7fff9dfcd668) at /home/kunlun/debugbuild/percona-server-8.0.18-9/sql/main.cc:25

Issue migrated from trac ticket # 154

component: kunlun-storage | priority: major | resolution: fixed

2021-09-06 19:19:45: @charleszettadb created the issue

CREATE OR REPLACE FUNCTION LOOP1()
RETURNS void
AS $$
DECLARE
i integer=1;
begin
for i in 1..6 LOOP
RAISE NOTICE '%',i;
END LOOP;
END;
$$ LANGUAGE PLPGSQL;
ERROR: syntax error at or near "1"
LINE 9: for i in 1..6 LOOP

2.                 ^
   

CREATE OR REPLACE FUNCTION LOOP1()
RETURNS void
AS $$
DECLARE
a integer=1;
begin
while a <= 6 loop
a = a+1
select n;
END LOOP;
END;
$$ LANGUAGE PLPGSQL;
ERROR: syntax error at or near "select"
LINE 9: select n;
^

Issue migrated from trac ticket # 339

component: kunlun-storage | priority: major

2021-12-21 12:52:49: @snowao created the issue

File_query_log::write_general() will not append '\n' to the tail of the output string when no extrainfo to be print.\
That is the reason why the test case failed.

Issue migrated from trac ticket # 335

component: kunlun-storage | priority: major

2021-12-20 15:13:36: @david-zhao created the issue

Issue migrated from trac ticket # 237

component: kunlun-storage | priority: major

2021-11-22 10:15:04: @david-zhao created the issue

In kunlun-storage-8.0.26, create the kunlun_sysdb database when installing a kunlun-storage instance, and create the heartbeat table and the below sequences table in kunlun_sysdb.

CREATE TABLE IF NOT EXISTS sequences( db varchar(512) not null, name varchar(512) not null, curval bigint not null, start bigint not null, step int not null, max_value bigint not null, min_value bigint not null, do_cycle bool not null, n_cache int unsigned not null, primary key(db,name) ) engine=innodb CHARACTER SET utf8 COLLATE utf8_general_ci STATS_PERSISTENT=0 comment='Kunlun sequence metadata used by computing nodes.' ROW_FORMAT=DYNAMIC

Issue migrated from trac ticket # 345

component: kunlun-storage | priority: major

2021-12-22 10:55:18: @snowao created the issue

Issue migrated from trac ticket # 140

component: kunlun-storage | priority: major | keywords: binlog;XA COMMIT ONE PHASE;rotation

2021-08-19 09:51:12: @david-zhao created the issue

This issue only exists in open source version of kunlun-storage.

If binlog rotation is initiated by a XA COMMIT ONE PHASE transaction commit, the worker thread blocks itself.

This is caused by two fixes which make sure XA COMMIT ONE PHASE txns are crash safe. Such fixes are correct but not sufficient --- we have to avoid doing rotation at XA COMMIT ONE PHASE commits.