А как насчет апгрейда серверов (по условию данного треда, т.е. в пределах одной версии но с UC N на FC M) работающих в паре репликации HADR ?
Казалось бы самое безопасное - остановить репликацию, прапгрейдить primary, заменить также бинарники secondary и восстановить репликацию обычным путем (level-0 на primary, физ. восстановление secondary, затем запуск репликации).
Я попробовал сделать на тестовой паре HADR апгрейд без отключения репликации (перед этим поменял SHMBASE в onconfig обоих серверов, поскольку они стали 64 bit). Остановил оба сервера, заменил бинарники версии 9.4 (AIX) UC1 на FC5 и запустил сервера. Репликация восстановилась !
Вот журналы сообщений в процессе запуска репликационных серверов.
SECONDARY LOG (начинал запуск с secondary, поскольку он был последним переведен в offline)
13:58:52 IBM Informix Dynamic Server Started.
...
13:58:57 IBM Informix Dynamic Server Version 9.40.FC5 Software Serial Number AAA#B000000
13:58:58 IBM Informix Dynamic Server Initialized -- Shared Memory Initialized.
13:58:58 DR: Logs already cleared
13:58:58 Physical Recovery Started at Page (1:443).
13:58:58 Physical Recovery Complete: 0 Pages Examined, 0 Pages Restored.
13:58:58 DR: Trying to connect to primary server = upgr_lo
13:58:58 DR: Cannot connect to primary server
13:58:58 DR: Turned off on secondary server
(в этот момент primary еще находится в offline)
13:58:58 Onconfig parameter SHMBASE modified from 805306368 to 504403158265495552.
13:58:59 Dataskip is now OFF for all dbspaces
13:58:59 Restartable Restore has been ENABLED
13:58:59 Recovery Mode
13:59:33 DR: Secondary server connected
(теперь primary тоже запущен, secondary его видит)
13:59:33 DR: Using default behavior of failure-recovering Secondary server
13:59:39 DR: Failure recovery from disk in progress ...
13:59:39 Logical Recovery Started.
13:59:39 10 recovery worker threads will be started.
13:59:39 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:59:39 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:59:39 Start Logical Recovery - Start Log 6, End Log ?
13:59:39 Starting Log Position - 6 0x8018
13:59:41 Checkpoint Completed: duration was 0 seconds.
13:59:41 Checkpoint loguniq 6, logpos 0x9018, timestamp: 0xd19b
13:59:41 Maximum server connections 0
13:59:43 Checkpoint Completed: duration was 0 seconds.
13:59:43 Checkpoint loguniq 6, logpos 0xbc924, timestamp: 0xf1f0
13:59:43 Maximum server connections 0
13:59:48 Checkpoint Completed: duration was 0 seconds.
13:59:48 Checkpoint loguniq 8, logpos 0x18c0c8, timestamp: 0x13639
13:59:48 Maximum server connections 0
13:59:49 DR: Secondary server operational
13:59:50 DR: Receiving index sysmaster:"informix".flags_text#flags_text_ix1 : Started
13:59:50 Checkpoint Completed: duration was 1 seconds.
13:59:50 Checkpoint loguniq 8, logpos 0x190018, timestamp: 0x1367b
13:59:50 Maximum server connections 0
13:59:51 DR: Receiving index sysmaster:"informix".flags_text#flags_text_ix1 : Completed.
14:01:45 Checkpoint Completed: duration was 0 seconds.
14:01:45 Checkpoint loguniq 9, logpos 0x149b8c, timestamp: 0x15bb6
14:01:45 Maximum server connections 0
14:01:46 DR: Receiving index sysutils:"informix".bar_server# 100_1 : Started
14:01:46 DR: Receiving index sysutils:"informix".bar_server# 100_1 : Completed.
14:01:47 DR: Receiving index sysutils:"informix".bar_object# 101_4 : Started
14:01:47 DR: Receiving index sysutils:"informix".bar_object# 101_4 : Completed.
14:01:59 DR: Receiving index sysutils:"informix".bar_object#bar_obj_idx : Started
14:01:59 DR: Receiving index sysutils:"informix".bar_object#bar_obj_idx : Completed.
14:02:01 DR: Receiving index sysutils:"informix".bar_action# 102_9 : Started
14:02:01 DR: Receiving index sysutils:"informix".bar_action# 102_9 : Completed.
14:02:02 DR: Receiving index sysutils:"informix".bar_action# 102_10 : Started
14:02:02 DR: Receiving index sysutils:"informix".bar_action# 102_10 : Completed.
14:02:04 DR: Receiving index sysutils:"informix".bar_action#bar_act_idx : Started
14:02:04 DR: Receiving index sysutils:"informix".bar_action#bar_act_idx : Completed.
14:02:06 DR: Receiving index sysutils:"informix".bar_instance# 103_14 : Started
14:02:06 DR: Receiving index sysutils:"informix".bar_instance# 103_14 : Completed.
14:07:07 Checkpoint Completed: duration was 1 seconds.
14:07:07 Checkpoint loguniq 10, logpos 0x11f018, timestamp: 0x19885
14:07:07 Maximum server connections 0
14:12:06 Checkpoint Completed: duration was 0 seconds.
14:12:06 Checkpoint loguniq 10, logpos 0x120018, timestamp: 0x1988b
14:12:06 Maximum server connections 0
PRIMARY LOG :
13:59:24 IBM Informix Dynamic Server Started.
...
13:59:30 IBM Informix Dynamic Server Version 9.40.FC5 Software Serial Number AAA#B000000
13:59:31 IBM Informix Dynamic Server Initialized -- Shared Memory Initialized.
13:59:31 DR: Reservation of the last logical log for log backup turned on
13:59:31 DR: Trying to connect to secondary server = upgr_p630
(primary увидел secondary)
13:59:33 DR: Primary server connected
13:59:33 DR: Using default behavior of failure-recovering Secondary server
13:59:33 Physical Recovery Started at Page (1:441).
13:59:33 Physical Recovery Complete: 0 Pages Examined, 0 Pages Restored.
13:59:33 Logical Recovery Started.
13:59:33 10 recovery worker threads will be started.
13:59:33 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:59:36 Logical Recovery has reached the transaction cleanup phase.
13:59:36 Logical Recovery Complete.
0 Committed, 0 Rolled Back, 0 Open, 0 Bad Locks
13:59:37 Onconfig parameter SHMBASE modified from 805306368 to 504403158265495552.
13:59:37 Dataskip is now OFF for all dbspaces
13:59:37 Checkpoint Completed: duration was 0 seconds.
13:59:37 Checkpoint loguniq 6, logpos 0x9018, timestamp: 0xcac6
13:59:37 Maximum server connections 0
13:59:37 On-Line Mode
13:59:37 Performing internal conversion of disk structures and databases...
Converting/Rebuilding 'sysmaster' database ...
13:59:39 Checkpoint Completed: duration was 0 seconds.
13:59:39 Checkpoint loguniq 6, logpos 0xbc924, timestamp: 0xe9d7
13:59:39 Maximum server connections 1
13:59:41 Logical Log 6 Complete, timestamp: 0xfdad.
13:59:41 DR: Sending log 6, size 500 pages, 53.00 percent used
13:59:44 Logical Log 7 Complete, timestamp: 0x13016.
13:59:44 DR: Sending log 7, size 500 pages, 67.20 percent used
13:59:47 Checkpoint Completed: duration was 1 seconds.
13:59:47 Checkpoint loguniq 8, logpos 0x18c0c8, timestamp: 0x134d3
13:59:47 Maximum server connections 1
13:59:47 DR: Sending log 8 (current), size 500 pages, 79.40 percent used
13:59:47 DR: Sending Logical Logs Completed
13:59:49 DR: Primary server operational
13:59:50 DR: Sending index sysmaster:"informix".flags_text#flags_text_ix1 : Started
13:59:50 Checkpoint Completed: duration was 0 seconds.
13:59:50 Checkpoint loguniq 8, logpos 0x190018, timestamp: 0x13520
13:59:50 Maximum server connections 1
13:59:50 Booting Language <spl> from module <>
13:59:50 Loading Module <SPLNULL>
(не установил бэкап логов в no_log.sh, пришлось забэкапить заполнившиеся журналы для продолжения процесса)
13:59:51 Logical Log Files are almost Full -- Backup is Needed.
In Data replication scenario, this could block failure-recovery
of the paired server.
13:59:51 Logical Log 8 Complete, timestamp: 0x13ffa.
13:59:51 DR: Sending index sysmaster:"informix".flags_text#flags_text_ix1 : Completed.
14:01:38 Logical Log 4 - Backup Started
14:01:38 Logical Log 4 - Backup Completed
14:01:39 Logical Log 5 - Backup Started
14:01:39 Logical Log 5 - Backup Completed
14:01:39 Logical Log 6 - Backup Started
14:01:39 Logical Log 6 - Backup Completed
14:01:39 Logical Log 7 - Backup Started
14:01:39 Logical Log 7 - Backup Completed
14:01:39 Logical Log 8 - Backup Started
14:01:40 Logical Log 8 - Backup Completed
14:01:41 Unloading Module <SPLNULL>
14:01:42 Loading Module <SPLNULL>
14:01:44 Checkpoint Completed: duration was 1 seconds.
14:01:44 Checkpoint loguniq 9, logpos 0x149b8c, timestamp: 0x154a3
14:01:44 Maximum server connections 1
14:01:45 Unloading Module <SPLNULL>
14:01:46 Logical Log 9 Complete, timestamp: 0x16cbf.
14:01:46 DR: Sending index sysutils:"informix".bar_server# 100_1 : Started
14:01:47 DR: Sending index sysutils:"informix".bar_server# 100_1 : Completed.
14:01:47 DR: Sending index sysutils:"informix".bar_object# 101_4 : Started
14:01:58 DR: Sending index sysutils:"informix".bar_object# 101_4 : Completed.
14:01:59 DR: Sending index sysutils:"informix".bar_object#bar_obj_idx : Started
14:02:00 DR: Sending index sysutils:"informix".bar_object#bar_obj_idx : Completed.
14:02:01 DR: Sending index sysutils:"informix".bar_action# 102_9 : Started
14:02:02 DR: Sending index sysutils:"informix".bar_action# 102_9 : Completed.
14:02:02 DR: Sending index sysutils:"informix".bar_action# 102_10 : Started
14:02:03 DR: Sending index sysutils:"informix".bar_action# 102_10 : Completed.
14:02:04 DR: Sending index sysutils:"informix".bar_action#bar_act_idx : Started
14:02:05 DR: Sending index sysutils:"informix".bar_action#bar_act_idx : Completed.
14:02:06 Loading Module <SPLNULL>
14:02:06 DR: Sending index sysutils:"informix".bar_instance# 103_14 : Started
14:02:06 Unloading Module <SPLNULL>
14:02:07 DR: Sending index sysutils:"informix".bar_instance# 103_14 : Completed.
14:02:07 'sysmaster' database built successfully.
14:02:07 Conversion Completed Successfully
14:07:06 Checkpoint Completed: duration was 1 seconds.
14:07:06 Checkpoint loguniq 10, logpos 0x11f018, timestamp: 0x19330
14:07:06 Maximum server connections 2
14:12:06 Checkpoint Completed: duration was 0 seconds.
14:12:06 Checkpoint loguniq 10, logpos 0x120018, timestamp: 0x19338
14:12:06 Maximum server connections 2
[
informix@h80lo /ism_clone/infxupgr $]onstat -g dri
IBM Informix Dynamic Server Version 9.40.FC5 -- On-Line (Prim) -- Up 00:29:48 -- 47424 Kbytes
Data Replication:
Type State Paired server Last DR CKPT (id/pg)
primary on upgr_p630 10 / 288
DRINTERVAL 30
DRTIMEOUT 30
DRLOSTFOUND /usr/informix/etc/dr.lostfound
[
informix@h80lo /ism_clone/infxupgr $]
Вот такой эксперимент.
