Всем привет !
Linux CentOS 5, Informix Dynamic Server Version 9.40.UC6
Натолкнулся на какую то фигню - После восстановления 0 уровня накатывается только один логический журнал ...
Все по порядку :
/opt/informix/etc/onconfig :
1.
2.
3.
4.
5.
6.
7.
8.
9.
ALARMPROGRAM /usr/scripts/bin/inf_alarm.sh # Alarm program path
TAPEDEV /backup/testcomp_base # Tape device path
TAPEBLK 8 # Tape block size (Kbytes)
TAPESIZE 999999999 # Maximum amount of data to put on tape (Kbytes)
LTAPEDEV /backup/testcomp_log # Log tape device path
LTAPEBLK 8 # Log tape block size (Kbytes)
LTAPESIZE 999999999 # Max amount of data to put on log tape (Kbytes)
Логи оттаскиваются скриптом /usr/scripts/bin/inf_alarm.sh который создает fifo и с одной стороны запускат ontape -a , а с другой cat fifo | gzip -c >testcomp_log.StartNumber-EndNumber.gz
Это бэкап 0 уровня и логи за ним :
1.
2.
3.
4.
5.
6.
7.
8.
9.
[root@testcomp 11]# l /backup/11/
total 178208
drwxrwxr-x 2 informix informix 4096 Mar 18 11:59 ./
drwxr-xr-x 7 informix informix 4096 Mar 18 11:53 ../
-rw-rw-r-- 1 informix informix 182232437 Mar 18 10:34 testcomp_base.0.gz.aa
-rw-rw---- 1 informix informix 7427 Mar 18 10:37 testcomp_log.255-255.gz
-rw-rw---- 1 informix informix 5434 Mar 18 10:40 testcomp_log.256-256.gz
-rw-rw---- 1 informix informix 893 Mar 18 10:41 testcomp_log.257-257.gz
-rw-rw---- 1 informix informix 4600 Mar 18 10:41 testcomp_log.258-258.gz
1. Разворачиваю бэкап 0 уровня :
1.
2.
3.
4.
5.
/etc/init.d/informix stop
FIFO=`grep '^TAPEDEV' $INFORMIXDIR/etc/onconfig | awk '{print $2}'`
mkfifo -m 660 $FIFO ; chown informix:informix $FIFO
for i in 1 2 ; do cat /backup/11/testcomp_base.0.gz.* | gzip -cd >$FIFO ; done &
ontape -r
...
Please mount tape 1 on /backup/testcomp_base and press Return to continue ... <ENTER>
...
Continue restore? (y/n)y
Do you want to back up the logs? (y/n)n
В логе :
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.
11:44:21 IBM Informix Dynamic Server Started.
11:44:21 Segment locked: addr=0x44000000, size=227794944
Wed Mar 18 11:44:22 2009
11:44:22 Event alarms enabled. ALARMPROG = '/usr/scripts/bin/inf_alarm.sh'
11:44:22 Booting Language <c> from module <>
11:44:22 Loading Module <CNULL>
11:44:22 Booting Language <builtin> from module <>
11:44:22 Loading Module <BUILTINNULL>
11:44:27 IBM Informix Dynamic Server Version 9.40.UC6
11:44:27 IBM Informix Dynamic Server Initialized -- Shared Memory Initialized.
11:44:27 Data replication type and state information reset. To start DR, use
the 'onmode -d' command and wait for the pair to be operational,
before shutting down the database server
11:44:27 Dataskip is now OFF for all dbspaces
11:44:27 Restartable Restore has been DISABLED
11:44:27 Recovery Mode
11:44:30 Physical Restore of rootdbs, physdbs, logicdbs, maindbs started.
11:44:30 Checkpoint Completed: duration was 0 seconds.
11:44:30 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b392b
11:44:30 Maximum server connections 0
11:44:30 Checkpoint Completed: duration was 0 seconds.
11:44:30 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b393b
11:44:30 Maximum server connections 0
11:44:30 Checkpoint Completed: duration was 0 seconds.
11:44:30 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b394b
11:44:30 Maximum server connections 0
11:46:15 Checkpoint Completed: duration was 0 seconds.
11:46:15 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b395a
11:46:15 Maximum server connections 0
11:46:15 Physical Restore of rootdbs, physdbs, logicdbs, maindbs Completed.
11:46:15 Checkpoint Completed: duration was 0 seconds.
11:46:15 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b3967
Restore a level 1 archive (y/n) n
Do you want to restore log tapes? (y/n)y
Roll forward should start with log number 255
Please mount tape 1 on /backup/testcomp_log and press Return to continue ...
В логе :
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
11:46:15 Maximum server connections 0
11:46:27 Logical Recovery Started.
11:46:27 1 recovery worker threads will be started.
11:46:27 Checkpoint Completed: duration was 0 seconds.
11:46:27 Checkpoint loguniq 255, logpos 0x12018, timestamp: 0x31b397c
11:46:27 Maximum server connections 0
11:46:27 Start Logical Recovery - Start Log 255, End Log ?
11:46:27 Starting Log Position - 255 0x12018
11:46:27 Clearing the physical and logical logs has started
11:46:39 Cleared 299 MB of the physical and logical logs in 12 seconds
Запускаю с другой консоли :
1.
2.
3.
LFIFO=`grep '^LTAPEDEV' $INFORMIXDIR/etc/onconfig | awk '{print $2}'`
mkfifo -m 660 $LFIFO ; chown informix:informix $LFIFO
cat /backup/11/testcomp_log* | gzip -cd >$LFIFO
Жму ENTER на сообщении
Please mount tape 1 on /backup/testcomp_log and press Return to continue ...
В логе :
1.
2.
3.
4.
11:53:26 Checkpoint Completed: duration was 0 seconds.
11:53:26 Checkpoint loguniq 255, logpos 0x13018, timestamp: 0x31b3989
11:53:26 Maximum server connections 0
Do you want to restore another log tape? (y/n)n
Далее onmode -m и вижу, что база в состоянии после лога 255 и до 256 ...
если же на предложение Do you want to restore another log tape? (y/n)
сказать 'y' и дать ему
1.
cat /backup/11/testcomp_log.256-256.gz | gzip -cd >$LFIFO
То он накатит 256 лог ... ну и т.д...
Так я помру догонять базу до актуала ...
Почему сконкатенированный поток из логов ( cat /backup/11/testcomp_log* | gzip -cd >$LFIFO ) воспринимается как один первый лог ?
В семерке тот же способ вроде работал правильно ... поток логов воспринимался нормально и догонялся правильно ...
P.S. Если сгрудить все логи в один простой файл
1.
2.
rm $LFIFO
cat /backup/11/testcomp_log* | gzip -cd >$LFIFO
И подставить его при восстановлении, ко картина таже ...
Может есть у кого какаие мысли ?
