Overview

This document is a guide for performing synchronization without service interruption for customers who cannot afford service downtime when using the TmaxTibero TSC product.

Test Environment

DB: T6

OS: CentOS 7.7

Configuration Information: 

 - PRIMARY Server: 192.168.41.250 (node 1)
 - STANDBY Server: 192.168.41.251 (node 2)
 - OBSERVER Server: 192.168.41.251 (node 2)

Method

Execution Test

1. Check OBSERVER User cm Status

$ su – OBSERVER → Connect to observer process user account
$ cmrctl show → Check observer cm status
Resource List of Observer OBSERVER
===================================================================
TSC_ID  CLS_ID   CM_NAME NID CM_STAT  INST_STAT PRI/TAR
-------- --------- ------------ ----- ---------- ------------ ---------------
   11         7      PRIMARY    1    UP         UP(NRML) PRIMARY
               6      STANDBY    1    UP         UP(RECO) TARGET
===================================================================
※ CM_NAME: Name set when registering cm resource (PRIMARY: node 1, STANDBY: node 2)
PRI/TAR: Current DB running status (PRIMARY: DB Normal running status, TARGET: DB Recovery running status)

2. Force Shutdown PRIMARY Server DB and Check OBSERVER cm Status

$ su – PRIMARY → Connect to DB user on PRIMARY running state
$ tbdown abnormal → Force shutdown DB
$ cmrctl show → Confirm service, db, vip stopped
Resource List of Node PRIMARY
=====================================================================
CLUSTER   TYPE    NAME         STATUS  DETAIL
----------- -------- -------------- -----------------------------------
COMMON network pub01        UP (public)  enp0s8
COMMON network net01         UP (private) 192.168.41.252/29000
COMMON  cluster  cls01          UP inc: net01, pub: pub01
    cls01          file cls01:0        UP /home/tsc1/tibero6/config/cls1_cfile
    cls01     service tibero         DOWN Database, Active Cluster (auto-restart: OFF)
    cls01          db tibero_p      DOWN tibero, /home/tsc1/tibero6, failed retry cnt: 0
    cls01         vip  vip01         DOWN tibero, 192.168.41.254/255.255.255.0/192.168.41.255 (1)
                                                    failed retry cnt: 0
=====================================================================
$ su – OBSERVER User
$ cmrctl show
Resource List of Observer OBSERVER
===================================================================
TSC_ID  CLS_ID  CM_NAME NID CM_STAT INST_STAT PRI/TAR
-------- -------- ----------- ----- --------- ----------- ---------
   11        7     PRIMARY    1       UP      DOWN       N
              8     STANDBY    1       UP      UP(NRML) PRIMARY
===================================================================
※ It can be confirmed that failover occurred to node 2 as node 1’s DB was forcibly shut down.

3. Recover PRIMARY Server and Start in Recovery Mode (Using suspend)

$ su – STANDBY → Connect to STANDBY user
$ cmrctl show → Check service, db, vip running
Resource List of Node STANDBY

=====================================================================
CLUSTER   TYPE    NAME         STATUS  DETAIL
----------- -------- -------------- -----------------------------------
COMMON network pub01        UP (public)  enp0s8
COMMON network net01         UP (private) 192.168.41.253/29000
COMMON  cluster  cls01          UP inc: net01, pub: pub01
    cls01          file cls01:0        UP /home/tsc1/tibero6/config/cls1_cfile
    cls01     service tibero         UP Database, Active Cluster (auto-restart: OFF)
    cls01          db tibero_s       UP(NRML) tibero, /home/tsc1/tibero6, failed retry cnt: 0
    cls01         vip  vip01         UP tibero, 192.168.41.254/255.255.255.0/192.168.41.255 (1)
                                          failed retry cnt: 0
=====================================================================
※ It can be confirmed that the service and vip are active.

4. Recover PRIMARY Server and Start in Recovery Mode (Using suspend)

$ su – STANDBY
$ tbsql sys/tibero
SQL> select name from v$datafile → Check DATAFILE paths
union all
select file_name from dba_temp_files;
SQL> alter database begin backup;
SQL> exit

$ sftp tibero@192.168.41.250 
sftp> cd /home/tsc1/tbdata
sftp> put /home/tsc1/tbdata/system001.dtf
sftp> put /home/tsc1/tbdata/undo001.dtf
sftp> put /home/tsc1/tbdata/usr01.dtf
sftp> put /home/tsc1/tbdata/tpr_ts.dtf
sftp> put /home/tsc1/tbdata/temp001.dtf
sftp> exit

$ tbsql sys/tibero
SQL> alter database end backup;
SQL> alter system suspend for standby;
SQL> exit

$ sftp tibero@192.168.41.250 
sftp> cd /home/tsc1/tbdata
sftp> put /home/tsc1/tbdata/redolog001.redo
sftp> put /home/tsc1/tbdata/redolog002.redo
sftp> put /home/tsc1/tbdata/redolog003.redo
sftp> put /home/tsc1/tbdata/c1.ctl
sftp> put /home/tsc1/tbdata/c2.ctl
sftp> exit
 

$ tbsql sys/tibero
SQL> alter system resume;
exit

※ If there are changes in archive files, copying archive files is also necessary
$ sftp tibero@192.168.41.250 
sftp> cd /home/tsc1/tbdata/tbarch
sftp> put log-t0-r0-s9.arc
sftp> put log-t0-r0-s10.arc
sftp> put log-t0-r0-s11.arc
 

$ su – PRIMARY → Connect to PRIMARY Server user
$ tbboot mount
$ tbsql sys/tibero
SQL> alter database standby controlfile;
SQL> alter database recover automatic for standby;
SQL> exit


$ tbdown immediate
$ tbboot recovery → Start PRIMARY Server as Target DB


$ su – OBSERVER User
$ cmrctl show
Resource List of Observer OBSERVER
===================================================================
TSC_ID CLS_ID   CM_NAME NID CM_STAT INST_STAT PRI/TAR
-------- -------- ----------- ----- --------- ----------- ---------
    11       7      PRIMARY   1       UP      UP(RECO) TARGET
              8      STANDBY   1       UP      UP(NRML) PRIMARY
===================================================================
※ Due to failover, the STANDBY (node 2) DB is running in Normal state, and the PRIMARY (node 1) DB is running in Recovery state as confirmed.