-
Notifications
You must be signed in to change notification settings - Fork 3
Expand file tree
/
Copy pathChangeLog
More file actions
5386 lines (3692 loc) · 177 KB
/
ChangeLog
File metadata and controls
5386 lines (3692 loc) · 177 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
2009-12-07 18:38 t-ishii
* NEWS, doc/pgpool-en.html, doc/pgpool-ja.html: Prepare 2.3
2009-12-07 17:11 t-ishii
* pool_process_query.c: Allow to load balance EXPLAIN and EXPLAIN
ANALYZE (only if query is SELECT). This will prevent kind mismatch
errors when query plan are different among servers.
2009-12-06 21:55 t-ishii
* main.c, pool.h, pool_type.h: Restore backend status by using
pgpool_status file.
2009-12-06 20:53 t-ishii
* version.h: Change code name for pgpool-II 2.3
2009-12-06 18:57 t-ishii
* pgpool.conf.sample-master-slave, pgpool.conf.sample-replication:
Add pgpool.conf sample for replication mode and master/slave mode
2009-12-06 17:46 t-ishii
* pgpool.conf.sample, pool.h, pool_config.c, pool_config.l,
pool_process_query.c, pool_process_reporting.c, pool_stream.c:
Add new configuration parameter fail_over_on_backend_error which
controls the behavior when write to backend socket fails.
Also add pool_log when postmaster administrative shutdown detected.
2009-12-06 16:52 t-ishii
* pool_process_query.c, pool_proto_modules.c: Ignore error of
pool_flush(frontend).
2009-12-06 16:46 t-ishii
* pool_proto_modules.c: Optimization when num_init_children == 1.
In this case we don't need to wait for master returning a response
before sending a query to the other nodes.
2009-12-06 16:43 t-ishii
* Makefile.am, Makefile.in, pool_process_query.c,
pool_process_reporting.c: Teach process_reporting() about
log_per_node_statement. Also refactor pool_process_query.c to
remove process_reporting() and make a new source file
"pool_process_reporting.c" to hold the funtcion. Main reason for
this is, process_reporting() is now big enough to have a separate C
source file.
2009-12-02 23:19 t-ishii
* child.c, pool.h, pool_process_query.c: Enhance SIGINT handler.
2009-12-02 23:11 t-ishii
* pool_proto_modules.c: Fix ProcessFrontendResponse() handling of
Parse message in master/slave mode. If it's inside transaction,
enable master_slave_dml, which means execute parse on master only.
This will prevent unwanted parase executed on slave, which holds at
least AccessShareLock on tables until transaction commits. It seems
unamed statement does not release the lock until execute.
Please note that, even with this fix, bind state will remain on
slave if JDBC issues parse with autocommit off. This is because
BEGIN parsed, bind on both master and slave but execute only on
master. However, this is not harmfull since parse/bind does not hold
any locks on tables.
Also add some logging statemets.
2009-12-01 22:05 t-ishii
* ChangeLog: v2.2.6
2009-12-01 22:02 t-ishii
* doc/pgpool-ja.html, NEWS: Add release note
2009-11-29 20:56 t-ishii
* pool_process_query.c: Fix is_select_query() not to allow cursor
statements. Close() should not allowed since hold cursor + update
may cause data inconsistency.
2009-11-29 17:42 t-ishii
* pool_process_query.c: Add logging by per_node_statement().
2009-11-29 17:29 t-ishii
* pool_proto_modules.c: Fix ProcessFrontendResponse() handling of
Parse message in master/slave mode. If it's inside transaction,
enable master_slave_dml, which means execute parse on master only.
This will prevent unwanted parase executed on slave, which holds at
least AccessShareLock on tables until transaction commits. It seems
unamed statement does not release the lock until execute.
Please note that, even with this fix, bind state will remain on
slave if JDBC issues parse with autocommit off. This is because
BEGIN parsed, bind on both master and slave but execute only on
master. However, this is not harmfull since parse/bind does not hold
any locks on tables.
2009-11-18 15:55 t-ishii
* config.h.in, configure, configure.in: Add -lcrypt for linking
commnad. Certain distribution ships libpq without linking to
libcrypt by some reasons which I don't know.
2009-11-18 15:45 t-ishii
* config.h.in, configure, configure.in: Add -lcrypt for linking
commnad. Certain distribution ships libpq without linking to
libcrypt by some reasons which I don't know.
2009-11-15 20:16 t-ishii
* ChangeLog: Run cvs2cl.pl -t --fsf -b -F trunk
2009-11-15 17:27 t-ishii
* pgpool.conf.sample, pool.h, pool_config.c, pool_config.l,
pool_process_query.c, pool_proto_modules.c, pool_relcache.c: Add
new directive log_per_node_statement. If true, print all statements
to the log. Similar to log_statement except that prints DB node id
and backend process id info.
2009-11-15 17:05 t-ishii
* pool_config.c, pool_config.l: Set change default value of
reset_query_list to "ABORT, DISCARD ALL"
2009-11-14 22:22 t-ishii
* child.c, pool.h, pool_process_query.c, pool_proto_modules.c,
pool_proto_modules.h: Fix connection_cache = false case. In this
mode, if frontend exits abnormaly, some global state variables are
not reset. For this purpose, new function reset_variables() added.
Remove unnecessary argument of end_load_balance().
2009-11-14 22:16 t-ishii
* pool.h, pool_process_query.c, pool_proto_modules.c,
pool_proto_modules.h, child.c: Fix connection_cache = false case.
In this mode, if frontend exits abnormaly, some global state
variables are not reset. For this purpose, new function
reset_variables() added.
Remove unnecessary argument of end_load_balance().
2009-11-14 20:41 t-ishii
* pool_config.c: Update pool_config.l
2009-11-14 20:40 t-ishii
* pool_config.l: Allow to change weight with reloading pgpool.conf
2009-11-14 20:38 t-ishii
* pool_config.c: Update pgpool_config.l
2009-11-14 20:37 t-ishii
* pool_config.l: Allow to change weight with reloading pgpool.conf
2009-11-14 20:35 t-ishii
* main.c: Limit listen(2)'s backlog parameter to
num_init_children*2 at largest. This is almost same algorithm as
PostgreSQL. I hope pgpool-II does not consume so much system
resource any more...
2009-11-14 08:35 t-ishii
* pool_timestamp.c, test/timestamp/expected/insert.out,
test/timestamp/expected/misc.out,
test/timestamp/expected/update.out: Fix some query rewriting
cases.
2009-11-12 16:37 t-ishii
* pgpool.conf.sample: Change reset_query_list default to PostgreSQL
8.3.
2009-11-12 16:04 t-ishii
* pool_process_query.c: Prepare for bind_rewrite_timestamp()
returns NULL case. This could happen if SELECT now() is called
inside an aborted transaction.
Fix SimpleForwardToBackend free rewrite_msg too early.
2009-11-11 10:53 t-ishii
* pool_timestamp.c: Fix comment.
2009-11-10 19:03 t-ishii
* Makefile.am, Makefile.in, pool_process_query.c,
pool_proto_modules.c, pool_proto_modules.h, pool_timestamp.c,
pool_timestamp.h, test/timestamp/Makefile, test/timestamp/main.c,
test/timestamp/parse_schedule, test/timestamp/run-test,
test/timestamp/expected/insert.out,
test/timestamp/expected/misc.out,
test/timestamp/expected/update.out,
test/timestamp/input/insert.sql, test/timestamp/input/misc.sql,
test/timestamp/input/update.sql: Add support for
CURRENT_TIMESTAMP and friends. Patch provided by Akio Ishida.
2009-11-10 11:24 t-ishii
* main.c: Fix comment.
2009-11-05 15:26 t-ishii
* pool_relcache.c: Fix relcache replacement logic bug. Patches
provided by Akio Ishida.
2009-11-05 13:12 t-ishii
* parser/outfuncs.c, test/parser/expected/insert.out,
test/parser/expected/select.out, test/parser/input/insert.sql,
test/parser/input/select.sql: Fix nodeToString() bug. example
test case: If nodeToString() is given the parse tree of "insert into
r1 values(1,2), (3,4)", it returns "insert into r1 values(1,2),
values(3,4)", which is apparently wrong. Patch provided by Akio
Ishida.
2009-11-05 12:43 t-ishii
* ChangeLog, TODO, parser/outfuncs.c,
test/parser/expected/insert.out, test/parser/expected/select.out,
test/parser/input/insert.sql, test/parser/input/select.sql: Fix
nodeToString() bug. example test case: If nodeToString() is given
the parse tree of "insert into r1 values(1,2), (3,4)", it returns
"insert into r1 values(1,2), values(3,4)", which is apparently
wrong. Patch provided by Akio Ishida.
2009-11-04 22:59 t-ishii
* pool_process_query.c: Fix do_query() to return more than 1
characters. Patch provided by Akio Ishida.
2009-11-03 20:14 devrim
* pgpool.spec:
- Remove init script from all runlevels before uninstall. Per #RH
Bugzilla 532177
2009-11-03 11:00 t-ishii
* pool.h, pool_auth.c, pool_process_query.c: When authentication
fails, report to frontend proper error messages, rather than
unhelpfull "server closed the connection unexpectedly" message.
Patch provided by Glyn Astill with minor tweak by Tatsuo Ishii.
2009-11-03 10:57 t-ishii
* pool.h, pool_auth.c, pool_process_query.c: When authentication
fails, report to frontend proper error messages, rather than
unhelpfull "server closed the connection unexpectedly" message.
Patch provided by Glyn Astill with minor tweak by Tatsuo Ishii.
2009-11-03 10:50 t-ishii
* pool_stream.c: Fix pool_flush_it() so that it does not complain
when failed to flush to frontend. Also add info to the message if it
was backend or frontend to avoid confusion.
2009-10-30 14:11 t-ishii
* pool_proto_modules.c: Allow is_temp_table to work with pre 8.4
PostgreSQL.
2009-10-30 14:01 t-ishii
* pool_proto_modules.c: Allow is_temp_table to work with pre 8.4
PostgreSQL.
2009-10-29 22:44 t-ishii
* Makefile.am, Makefile.in, child.c, configure, configure.in,
pool.h, pool_process_query.c, pool_proto_modules.c: Add
pool_relcache.c to Makefile.am. Refactor need_insert_lock to use
pool_search_relcache and friends. Add LocalSessionId to child.c
which manage per session counter of relcache. Allow to use temp
tables with master/slave mode. Parse() failed. since Parse() was
always sent to master and slave and slave does not have the temp
table. To fix this, is_temp_table() added to pool_proto_modules.c
which inquire particular table is a temporary one or not when
INSERT/UPDATE/DELETE. If it's a temporary table, parse message is
sent to only master. Note that /*NO LOAD BALANCE*/ directive can be
used to control this. SELECT still needs /*NO LOAD BALANCE*/. It's
hard to find all tables in a SELECT statement.
2009-10-29 21:58 t-ishii
* child.c, pool.h, pool_process_query.c, Makefile.am, Makefile.in,
configure, configure.in, pool_proto_modules.c: Add
pool_relcache.c to Makefile.am. Refactor need_insert_lock to use
pool_search_relcache and friends. Add LocalSessionId to child.c
which manage per session counter of relcache. Allow to use temp
tables with master/slave mode. Parse() failed. since Parse() was
always sent to master and slave and slave does not have the temp
table. To fix this, is_temp_table() added to pool_proto_modules.c
which inquire particular table is a temporary one or not when
INSERT/UPDATE/DELETE. If it's a temporary table, parse message is
sent to only master. Note that /*NO LOAD BALANCE*/ directive can be
used to control this. SELECT still needs /*NO LOAD BALANCE*/. It's
hard to find all tables in a SELECT statement.
2009-10-29 21:46 t-ishii
* pool_relcache.c: Per process relation cache module
2009-10-29 21:42 t-ishii
* pool_relcache.c: Per process relation cache module
2009-10-25 16:19 t-ishii
* test/jdbc/: CreateTempTableTest.java, expected/CreateTempTable,
RunTest.java: Add temp table test
2009-10-25 16:14 t-ishii
* test/jdbc/: CreateTempTableTest.java, expected/CreateTempTable:
Add temp table test
2009-10-22 20:52 devrim
* redhat/: pgpool.init, pgpool.sysconfig: Fix logging, per report
from Toshihiro Kitagawa.
2009-10-21 23:39 t-ishii
* pool_process_query.c: Do not execute CREATE/DROP TABLE SPACE
inside a transaction block.
Teach do_command() that S (parameter status) should be ignored until
receiving command complete.
2009-10-21 23:05 t-ishii
* main.c: Let reaper() report the case when child died by segfault.
2009-10-12 11:21 t-ishii
* pgpool.8.in: Fix typo in man page. Patch provided by Rodolphe.
2009-10-12 11:19 t-ishii
* pgpool.8.in: Fix typo in man pages. Patch provided by Rodolphe.
2009-10-12 10:54 t-ishii
* pool_process_query.c: REINDEX DATABASE or SYSTEM should not be
executed in a transaction block. See [pgpool-general-jp: 623] for
more details.
2009-10-12 10:41 t-ishii
* pool_process_query.c: REINDEX DATABASE or SYSTEM should not be
executed in a transaction block. See [pgpool-general-jp: 623] for
more details.
2009-10-06 00:13 devrim
* pgpool.spec:
- Update to 2.2.5, for various fixes described at
http://lists.pgfoundry.org/pipermail/pgpool-general/2009-October/002188.html
- Re-apply a fix for Red Hat Bugzilla #442372
- Add 2 new docs, per Tatsuo.
2009-10-06 00:12 devrim
* pgpool.spec:
Add 2 new docs, per Tatsuo.
2009-10-05 19:38 devrim
* pgpool.spec:
- Update to 2.2.5, for various fixes described at
http://lists.pgfoundry.org/pipermail/pgpool-general/2009-October/002188.html
- Re-apply a fix for Red Hat Bugzilla #442372
- Sync with Fedora spec.
2009-10-04 18:15 t-ishii
* NEWS, configure.in, configure, ChangeLog (utags: V2_2_5): Prepare
2.2.5
2009-10-04 18:15 t-ishii
* doc/pgpool-ja.html (tags: V2_2_5): Add release note for 2.2.5
2009-10-04 08:52 t-ishii
* doc/: tutorial-en.html (tags: V2_2_5), tutorial-en.html,
tutorial-ja.html: Add sample to use rotatelogs
2009-10-03 21:45 t-ishii
* doc/tutorial-ja.html (tags: V2_2_5): Add sample to use rotatelogs
2009-10-03 21:39 t-ishii
* sample/: pgpool_recovery_pitr (tags: V2_2_5),
pgpool_recovery_pitr: Enhance online recovery 2nd stage script to
sync sequence value
2009-10-03 20:55 t-ishii
* pool_proto_modules.h (tags: V2_2_5), pool_proto_modules.h: Remove
unnecessary declaration "allow_close_transaction"
2009-10-02 17:00 t-ishii
* pool_proto_modules.c: Prevoius commit was incomplete
2009-10-02 16:56 t-ishii
* pool_proto_modules.c (tags: V2_2_5), pool_proto_modules.c: Fix
Parse() to use wait_for_query_response() instead of synchronize().
2009-10-02 16:53 t-ishii
* pool_process_query.c (tags: V2_2_5), pool_process_query.c: Fix
comment
2009-09-26 18:14 t-ishii
* pool_proto_modules.c: Allow to print more accurate query when
Parse() detects kind mismatch errors.
2009-09-26 16:51 t-ishii
* pool_process_query.c: Fix reset_backend to deal with DEALLOCATE
error. In this case prepared objects are not removed and
reset_backend is infinitely called.
Fix do_command to guard against the case that after execution of an
SQL command, the backend does not return either N, E or C. This
should not happen but there was a report that it actually happend in
certain cases(maybe caused by asynchronized backend data?)
See:
Subject: Re: [Pgpool-general] pgpool 2.2.4: DEALLOCATED children
From: Xavier Noguer Date: Fri, 25 Sep 2009 03:35:57 -0400
For more details.
Patches by Xavier Noguer with minor tweak by Tatsuo Ishii.
2009-09-26 16:36 t-ishii
* pool_process_query.c: Fix reset_backend to deal with DEALLOCATE
error. In this case prepared objects are not removed and
reset_backend is infinitely called.
Fix do_command to guard against the case that after execution of an
SQL command, the backend does not return either N, E or C. This
should not happen but there was a report that it actually happend in
certain cases(maybe caused by asynchronized backend data?)
See:
Subject: Re: [Pgpool-general] pgpool 2.2.4: DEALLOCATED children
From: Xavier Noguer Date: Fri, 25 Sep 2009 03:35:57 -0400
For more details.
Patches by Xavier Noguer with minor tweak by Tatsuo Ishii.
2009-09-26 09:27 t-ishii
* pool_process_query.c: Remove error checks for
pool_write(frontend), pool_flush(frontend) in
SimpleForwardToFrontend.
Check error for send_deallocate() in reset_backend. Otherwise
send_deallocate will be called infinitely. Per bug report from
Agustin Almonte Ferrada. See:
Subject: Re: [Pgpool-general] pgpool 2.2.4: DEALLOCATED children
From: Agustin Almonte Ferrada <aalmonte@antica.cl> To: Tatsuo Ishii
<ishii@sraoss.co.jp> Cc: pgpool-general@pgfoundry.org Date: Wed, 23
Sep 2009 01:55:54 -0400
2009-09-26 09:05 t-ishii
* pool_process_query.c: FIX parallel query
1. Fix JOIN Expr
Subquery was used on larg or rarg.
Subquery was used on ON-Cluase.
Full OUTER JOIN was not optimized.
2. Fail_on_error=false was passed to the 3rd arguments of
dblink.
This patch was applied to only 2.2-STABLE tree by y-mori on
2009/09/08.
2009-09-25 23:36 t-ishii
* pool_proto_modules.c: Make allow_close_transaction to static and
add comments.
Let execute CREATAE TEMP TABLE on every nodes in master/slave mode.
If not, subsequent Parse or Bind to use the table will fail. At
this point this doen't work since MASTER_SLAVE set to 0 in
ProcessFrontendResponse. More analysis is needed...
2009-09-25 16:35 t-ishii
* main.c (tags: V2_2_5), main.c: Fix write_pid_file. It forgot to
write out NULL termination of pid. Pgpool occasionally won't stop
because of this bug.
2009-09-23 20:39 t-ishii
* pool_process_query.c: Remove error checks for
pool_write(frontend), pool_flush(frontend) in
SimpleForwardToFrontend.
Check error for send_deallocate() in reset_backend. Otherwise
send_deallocate will be called infinitely. Per bug report from
Agustin Almonte Ferrada. See:
Subject: Re: [Pgpool-general] pgpool 2.2.4: DEALLOCATED children
From: Agustin Almonte Ferrada <aalmonte@antica.cl> To: Tatsuo Ishii
<ishii@sraoss.co.jp> Cc: pgpool-general@pgfoundry.org Date: Wed, 23
Sep 2009 01:55:54 -0400
2009-09-23 11:03 t-ishii
* pool_auth.c (tags: V2_2_5), pool_auth.c: Fix pool_read_kind and
pool_read_int so that the error message for kind is shown in
hexadecimal. Other part of pgpool-II already does this.
2009-09-17 20:32 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h:
Teach do_command() to use wait_for_query_response(). insert_lock()
calls do_command() to issue LOCK TABLE to lock tables having SERIAL
column when insert_lock is true. Problem is, insert_lock() waits
forever in do_command() if the target table is locked by someone
else. This patch should give a way insert_lock() to espcape from the
situation.
2009-09-17 20:17 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h:
Teach do_command() to use wait_for_query_response(). insert_lock()
calls do_command() to issue LOCK TABLE to lock tables having SERIAL
column when insert_lock is true. Problem is, insert_lock() waits
forever in do_command() if the target table is locked by someone
else. This patch should give a way insert_lock() to espcape from the
situation.
2009-09-06 12:56 t-ishii
* child.c, pool.h: Fix connection_count_down() so that it does
decrement too much in some corner cases. If failed to read a start
up packet, or receive cancel request etc., connection_count_down()
is called and goes back to the onnection accept loop. Problem is, at
the very beginning of the connection accept loop, if we have
received a signal, we call child_exit() which calls
connection_count_down() again. Problem report by Ryoichi TANABE.
See [pgpool-general-jp: 616] for more details.
Also unneccesary public declarations for connection_count_up() and
connection_count_down() are removed.
2009-09-06 12:54 t-ishii
* pool_process_query.c: Undef #define SEND_NOTICE_ON_PROTO2, which
was defined by mistake.
2009-09-06 12:52 t-ishii
* child.c, pool.h (utags: V2_2_5): Fix connection_count_down() so
that it does decrement too much in some corner cases. If failed to
read a start up packet, or receive cancel request etc.,
connection_count_down() is called and goes back to the onnection
accept loop. Problem is, at the very beginning of the connection
accept loop, if we have received a signal, we call child_exit()
which calls connection_count_down() again. Problem report by
Ryoichi TANABE. See [pgpool-general-jp: 616] for more details.
Also unneccesary public declarations for connection_count_up() and
connection_count_down() are removed.
2009-08-24 13:58 t-ishii
* ChangeLog (tags: V2_2_4): Prepare 2.2.4
2009-08-24 13:55 t-ishii
* NEWS, configure, doc/pgpool-ja.html (utags: V2_2_4): Prepare
2.2.4
2009-08-24 13:53 t-ishii
* configure.in (tags: V2_2_4): Prepare 2.2.4.
2009-08-24 13:36 t-ishii
* TODO (tags: V2_2_5, V2_2_4): Add a few items
2009-08-22 13:19 t-ishii
* child.c, main.c, md5.c (tags: V2_2_5), pcp_child.c (tags:
V2_2_5), pg_md5.c (tags: V2_2_5), pool_auth.c, pool_config.c
(tags: V2_2_5), pool_connection_pool.c (tags: V2_2_5),
pool_error.c (tags: V2_2_5), pool_hba.c (tags: V2_2_5), pool_ip.c
(tags: V2_2_5), pool_params.c (tags: V2_2_5), pool_path.c (tags:
V2_2_5), pool_process_query.c, pool_proto_modules.c,
pool_query_cache.c (tags: V2_2_5), pool_rewrite_outfuncs.c (tags:
V2_2_5), pool_rewrite_query.c (tags: V2_2_5), pool_sema.c (tags:
V2_2_5), pool_shmem.c (tags: V2_2_5), pool_signal.c (tags:
V2_2_5), pool_stream.c (tags: V2_2_5), pool_system.c (tags:
V2_2_5), ps_status.c (tags: V2_2_5), recovery.c (tags: V2_2_5),
strlcpy.c (tags: V2_2_5), pcp/md5.c (tags: V2_2_5) (utags:
V2_2_4): Remove unnecessary spaces and tabs at end of line.
Patch conributed by Jun Kuriyama.
2009-08-22 13:04 t-ishii
* child.c, main.c, md5.c, pcp_child.c, pg_md5.c, pool_auth.c,
pool_config.c, pool_connection_pool.c, pool_error.c, pool_hba.c,
pool_ip.c, pool_params.c, pool_path.c, pool_process_query.c,
pool_proto_modules.c, pool_query_cache.c,
pool_rewrite_outfuncs.c, pool_rewrite_query.c, pool_sema.c,
pool_shmem.c, pool_signal.c, pool_stream.c, pool_system.c,
ps_status.c, recovery.c, strlcpy.c, pcp/md5.c: Remove unnecessary
spaces and tabs at end of line. Patch conributed by Jun Kuriyama.
2009-08-20 15:26 t-ishii
* pool_process_query.c: Fix possible bug introduced in pgpool-II
2.2.2. Load balance control variables may remain not be restored
and subsequent DML/DDL call might sent to only master node.
2009-08-18 17:41 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h:
Change wait_for_query_response() so that it sends NOTICE message to
frontend periodically. It appears that pgpool-II 2.2.3
unconditionaly sends param packet to client even it uses version 2
protocol, which is apparentlt wrong. Also tweak checking period from
1 second to 30 seconds since 1 second seems too aggressive. Note
that you need to edit pool_process_query.c to define
SEND_NOTICE_ON_PROTO2 around line 879 to enable this feature in
version 2 protocol.
2009-08-18 17:38 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h
(tags: V2_2_4): Change wait_for_query_response() so that it sends
NOTICE message to frontend periodically. It appears that pgpool-II
2.2.3 unconditionaly sends param packet to client even it uses
version 2 protocol, which is apparentlt wrong. Also tweak checking
period from 1 second to 30 seconds since 1 second seems too
aggressive.
2009-08-18 17:30 t-ishii
* child.c: Remove unneccessary call to send_frontend_exit() in
do_child().
2009-08-13 22:40 t-ishii
* child.c: Remove unnessary call of send_frontend_exits() in die().
2009-08-13 22:39 t-ishii
* main.c: Block signals before forking children rather after.
Otherwise parent will be killed by failover signal if it receives
before establish signal handler. See [pgpool-general-jp: 586] for
more details.
2009-08-13 22:36 t-ishii
* child.c: Remove unnessary call of send_frontend_exits() in die().
2009-08-13 22:34 t-ishii
* main.c: Block signals before forking children rather after.
Otherwise parent will be killed by failover signal if it receives
before establish signal handler. See [pgpool-general-jp: 586] for
more details.
2009-08-11 22:53 t-ishii
* ChangeLog (tags: V2_2_3): Prepare 2.2.3.
2009-08-11 22:52 t-ishii
* configure.in, configure (utags: V2_2_3): Stamp 2.2.3.
2009-08-11 22:48 t-ishii
* doc/pgpool-ja.html (tags: V2_2_3): Fix description of parallel
query. Add release note for 2.2.3.
2009-08-11 22:23 t-ishii
* pool_connection_pool.c (tags: V2_2_3), pool_connection_pool.c:
Fix new_connection() so that it does leave garbage in
POOL_CONNECTION_POOL->slot[i]. This caused segfault in child_exit().
Bug report at [pgpool-general-jp: 586].
2009-08-07 14:27 y-mori
* pool_process_query.c, pool_rewrite_outfuncs.c (utags: V2_2_3):
FIX parallel query
1. Fix JOIN Expr
Subquery was used on larg or rarg.
Subquery was used on ON-Cluase.
Full OUTER JOIN was not optimized.
2. Fail_on_error=false was passed to the 3rd arguments of dblink.
2009-08-06 20:38 t-ishii
* Makefile.in: Update Makefile.am
2009-08-06 18:11 t-ishii
* Makefile.am: Add parser/test files. Patch contributed by Akio
Ishida.
2009-08-06 16:55 t-ishii
* pool_proto_modules.c (tags: V2_2_3), pool_proto_modules.c: Print
last query for kind mismatch error message even extended protocol is
used. Patch contributed by Akio Ishida.
2009-08-06 16:46 t-ishii
* pool_process_query.c: Fix buffer overrun error which mangled kind
mismatch error message in pool_send_error_message(). Bug rpeort by
Takahiro Itagaki at [pgsql-jp: 39950]. Fix suggested by Akio Ishida
at [pgsql-jp: 39952].
2009-08-06 16:38 t-ishii
* pool_process_query.c: Fix buffer overrun error which mangled kind
mismatch error message in pool_send_error_message(). Bug rpeort by
Takahiro Itagaki at [pgsql-jp: 39950]. Fix suggested by Akio Ishida
at [pgsql-jp: 39952].
2009-08-03 19:31 t-ishii
* parser/: kwlist.h, kwlookup.c: Missing files for PostgreSQL 8.4
paraser adaoptation. Patch contributed by Akio Ishida.
2009-08-01 20:41 t-ishii
* child.c: Check MASTER_CONNECTION(p)->sp) is non 0 before further
referring to sp. Current implementation may interrupt in the window,
after setting p and before setting sp.
2009-08-01 20:40 t-ishii
* child.c (tags: V2_2_3): Check MASTER_CONNECTION(p)->sp) is non 0
before further referring to sp. Current implementation may interrupt
in the window, after setting p and before setting sp. Per bug report
from Jun Kuwamura.
2009-08-01 20:34 t-ishii
* pool_connection_pool.c: Fix pool_debug's mismatch between data
type and conversion specification for time_t. Check
MASTER_CONNECTION(p)->sp) is non 0 before further referring to sp.
Current implementation may interrupt in the window, after setting p
and before setting sp.
2009-08-01 12:20 t-ishii
* test/parser/: parse_schedule, input/v84.sql: parser test for
PostgreSQL 8.4 or later. Patch contributed by Akio Ishida.
2009-07-30 23:44 t-ishii
* pool_process_query.c, pool_rewrite_outfuncs.c, pool_type.h,
parser/Makefile.am, parser/Makefile.in, parser/copyfuncs.c,
parser/gram.c, parser/gram.h, parser/gram.y, parser/keywords.c,
parser/keywords.h, parser/makefuncs.c, parser/makefuncs.h,
parser/nodes.c, parser/nodes.h, parser/outfuncs.c,
parser/parsenodes.h, parser/parser.c, parser/pool_memory.h,
parser/pool_parser.h, parser/primnodes.h, test/parser/Makefile,
test/parser/expected/copy.out, test/parser/expected/create.out,
test/parser/expected/cursor.out, test/parser/expected/delete.out,
test/parser/expected/insert.out, test/parser/expected/misc.out,
test/parser/expected/prepare.out,
test/parser/expected/privileges.out,
test/parser/expected/select.out, test/parser/expected/update.out:
Adopt PostgreSQL 8.4 parser. Patch contributed by Akio Ishida.
2009-07-30 16:46 t-ishii
* pool_process_query.c: Fix typo in comment
2009-07-28 14:52 y-mori
* doc/: pgpool-en.html (tags: V2_2_5, V2_2_4, V2_2_3),
pgpool-ja.html: add restrictions for parallel mode
-Natural Join
=> not support
-USING CLAUSE at Join Expression
=> auto convert to ON CLAUSE
2009-07-28 09:34 t-ishii
* pool_process_query.c: Fix typo in comment
2009-07-26 18:14 t-ishii
* pool_rewrite_outfuncs.c: Fix breakage by previous commit. Patch
by Guillaume Lelarge.
2009-07-26 18:04 t-ishii
* pool_rewrite_outfuncs.c: Fix breakage by previous commit. Patch
by Guillaume Lelarge.
2009-07-26 12:09 t-ishii
* pool_rewrite_outfuncs.c: fix: Join statment in parallel query
sample)
Accouts table is defined by dist_def,
the below query is executed as loadbalance.
SELECT a.aid
FROM
accounts as a
JOIN
accounts as b
ON a.aid=b.aid
WHERE a.aid =1 or a.aid=2;
2009-07-26 12:01 t-ishii
* pool_process_query.c: Fix comment about load balancing.
2009-07-26 11:57 t-ishii
* pool_sema.c (tags: V2_2_3), pool_sema.c: Fix various
pool_error/log format errors. Patch provided by Akio Ishida.
2009-07-24 21:30 y-mori
* pool_rewrite_outfuncs.c: fix: Join statment in parallel query
sample)
Accouts table is defined by dist_def,
the below query is executed as loadbalance.
SELECT a.aid
FROM
accounts as a
JOIN
accounts as b
ON a.aid=b.aid
WHERE a.aid =1 or a.aid=2;
2009-07-22 17:51 t-ishii
* child.c, main.c, pool.h, pool_process_query.c,
pool_proto_modules.c, pool_rewrite_outfuncs.c, pool_shmem.c: Fix
various pool_error/log format errors. Patch provided by Akio
Ishida.
2009-07-22 17:46 t-ishii
* child.c, main.c (tags: V2_2_3), pool.h (tags: V2_2_4, V2_2_3),
pool_process_query.c, pool_proto_modules.c,
pool_rewrite_outfuncs.c, pool_shmem.c (tags: V2_2_3): Fix various
pool_error/log format errors. Patch provided by Akio Ishida.
2009-07-22 17:26 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h:
Query cancel by statement timeout does not generate "kind mismatch"
error any more. How to reproduce the test case:
Directory connect to PostgreSQL. T1: BEGIN; T1: LOCK TABLE t1;
T2: SET statement_timeout TO 5000; T2: UPDATE t1 SET ...
UPDATE query aborts and error message comes out on T2: ERROR:
canceling statement due to statement timeout
A side effect of this is, statement timeout works correctly (rather
than statement timeout value * # of node).
Note that to initiate T1 by directory connecting to slave node will
not work (still results in kind mismatch error). But I don't think
this could happend in the real world.
Also refactor error detecting modules (detect_deadlock_error and so
on).
2009-07-22 14:33 t-ishii
* pool_process_query.c, pool_proto_modules.c, pool_proto_modules.h
(tags: V2_2_3): Query cancel by statement timeout does not
generate "kind mismatch" error any more. How to reproduce the test
case:
Directory connect to PostgreSQL. T1: BEGIN; T1: LOCK TABLE t1;
T2: SET statement_timeout TO 5000; T2: UPDATE t1 SET ...
UPDATE query aborts and error message comes out on T2: ERROR:
canceling statement due to statement timeout
A side effect of this is, statement timeout works correctly (rather
than statement timeout value * # of node).
Note that to initiate T1 by directory connecting to slave node will
not work (still results in kind mismatch error). But I don't think
this could happend in the real world.
Also refactor error detecting modules (detect_deadlock_error and so
on).
2009-07-21 18:48 t-ishii
* main.c: Check read(2) error in health_check(). This should detect
case described in message below, i.e. postmaster is blocked by
SIGSTOP.
Subject: [Pgpool-general] Healt Check issues From: Vctrjsn@gmail.com
To: pgpool-general@pgfoundry.org Date: Sun, 05 Jul 2009 18:01:18
+0000
2009-07-21 12:59 t-ishii
* ChangeLog: Update Change. We are getting closer to 2.2.3.
2009-07-21 10:43 t-ishii
* pool.h, pool_proto_modules.c: Check frontend connection while
waiting for response from backend for extended protocol (Execute()).
Replace synchronize() with wait_for_query_response(). Also tweak
pool.h so that it includes cancel packet's mazic proto number.
2009-07-21 10:36 t-ishii
* pool.h, pool_proto_modules.c: Check frontend connection while
waiting for response from backend for extended protocol (Execute()).
Replace sychronize() with wait_for_query_response(). Also tweak
pool.h so that it includes cancel packet's mazic proto number.
2009-07-21 09:30 t-ishii
* pool_process_query.c: Do not start transaction if the command is
CLUSTER with no argument
2009-07-21 09:26 t-ishii
* pool_process_query.c: Do not start transaction if the command is
CLUSTER with no argument
2009-07-20 21:21 t-ishii
* child.c, pool.h, pool_process_query.c, pool_proto_modules.c,
pool_proto_modules.h: Check frontend connection while waiting for
response from backend. This is done in wait_for_query_response() by
writing a dummy parameter status packet to pgpool client every 1
second. If the write fails, a cancel packet is sent to backend to
free backend. This is usefull to prevent pgpool child from waiting
forever for backend response. Pgpool client will diconnect to pgpool
if the query takes too long. This is typical use case for Web
applications.
2009-07-20 21:16 t-ishii
* pool_process_query.c: Fix bug with SimpleForwardToFrontend().
Clients using the extend protocol send 'H'(Flush) after each
command. Problem is, SimpleForwardToFrontend does not immediately
forward 'C'(Command Complete) packet to frontend. So the client
sends 'H' and wait for 'C'. On the other hand backend receives 'H'
but reply back nothing since Command complete has nothing to be
returned. Both side stuck.