Create Standby Database (CSD)
- Jan Klinke
Dbvisit has guided process which will guide you through the process of creating (instantiating) standby database. CSD process includes backup, tansfer and restore of primary database to the standby server. CSD process can be run in dbvcontrol, dbvctl command line, dbvcli command line and API. This page, will show you how to use dbvcontrol to run CSD process.
Duration of CSD process is highly dependent on specific system and it’s difficult to predict. Good starting point is how long typical full RMAN backup and RMAN restore of your production database takes.
1. Important Considerations
You need to fulfill allPrerequisites and Prepare Standby Database File Locations before initiating this process
Backup transfer and restore are done per single datafile and is parallelized. For example restore of datafile #2 runs in parallel to backup of datafile #3.
it is not possible to use or provide existing backup
by default, backup is compressed. If you have enough disk space and fast network, you can make backup and restore faster by specifying following variable in DDC file:
RMAN_DBF_BACKUP_TYPE=AS BACKUPSETSee <TBD> how to modify DDC file.
2. Initiating the CSD process
You can initiate the process by selecting “Setup Up Now”:
or running “Setup Disaster Recovery” from the ACTIONS pane:
3. CSD process options
There are several options you will need to consider when running CSD:
Number | Description | Explanation |
|---|---|---|
1 | Do you want to ship archive logs from the Primary database while creating the Standby database? | By default, Dbvisit will start archivelog shipping from the moment the CSD process starts, which then helps for standby database quick synchronization after the backup is restored. There’s usually no reason to disable this option. |
2 | Do you want to perform backup, transfer and restore operations in parallel when possible? | By default, Dbvisit will parallelize backup, transfer and restore, which makes the whole CSD process faster. There’s usually no reason to disable this option. Note this isn’t related to Oracle Enterprise Edition parallelism at all. |
3 | Is the Standby backup destination transportable media? | Enable only if you want to manually perform the transfer of backup files to standby server (for example portable disk). This procedure is described in more detail further on this page in separate chapter |
4 | Do you want to use SHARED SPACE to backup the database? | You have the option to use shared space between primary and standby server (for example provided by NFS). In this case, network transfer will be skipped. |
5 | Enter the Source and Target temp folder locations. | Enter backup folders on primary and standby which will be able to hold primary database backup. Backup is by default compressed. On primary server only, the backup of datafile is deleted immediately after it is transferred to standby. Do not user drive root locations without a subfolder. For example, don’t use C:\ or /u01 mounts directly. Always use a subfolder like: C:\temp, /u01/backup) |
After ensuring all settings are correct, proceed to next part where Standby Database initialization parameters can be adjusted.
If Dbvisit will detects that there’s a oracle grid infrastructure on standby server (for example when standby database will be placed on ASM), following option will be automatically included:
Enabled (Selecting Yes): standby database will be registered with srvctl, any existing srvctl resource with matching db_unique_name will be removed
Disabled: any existing srvctl resource with matching db_unique_name will be removed
4. Adjusting standby database init parameters
If you are doing any path conversion (including ASM to filesystem) you must modify appropriate paramters and you can’t skip this section. Refer to this page:
for correct parameter settings.
CSD process allows you to set & modify standby database init parameters. This is useful when you want to for example change memory allocation or use different paths for standby database datafiles. Initially the parameters are copy of primary database parameters.
Adding or modification of init parameters is easy, you first need to expand the complete list (1):
Then you can add new parameter (2), edit existing (3), or delete it completely (4).
When entering stings, you don’t have to use single quotes. Also make sure to avoid any trailing slashes when entering filesystem paths (i.e.: don’t use
/u02/app/oracle/oradata/)
5. Final Settings and starting the CSD process
In the last part, you can choose wheter to automatically start sychronization in the background and set its interval (1), you can enable observer(2) and save all you customization of this form page as template (3) - it will be available for future use.
We recommend to leave the automated standby update and observer to “Enabled”
Submitting the form “Create Standby Database” will start the CSD process. CSD process is fully automated and there’s no further intervention required.
You can check the CSD process status by viewing the task status:
All 10 steps must be completed successfully. You can monitor the progress of backup & restore by checking backup directories you specified on primary and standby. If parallelism wasn’t disabled, you will be able to see standby instance running and datafiles restored shortly after backup of datafile 1 is restored.
5. Restarting the CSD process
It can happen that the create standby database process will fail due to network disconnection or disk full, for example:
In that case, it’s possible restart create standby database process wihtout starting from scratch after the root cause of the failure is fixed. The steps are following:
Stop manually standby instance
export ORACLE_SID=<your_standby_oracle_sid>
sqlplus / as sysdba
SQL> shu abort;Once Standby Database instance completes, refresh DASHBOARTD and choose “Resume Creation”:
or select ACTIONS and choose same option:
On the next page, confirm the resuming of the CSD:
Create standby database process will then restart from the last successfult step (backup/restore/transfer of datafile X).
6. CSD With Transportable Media option
Transportable Media option is provided to enable you to easily create standby database in remote site with poor network bandwidth. The steps are:
dbvisit creates backup on primary and pauses CSD process
user is prompted to manually transfer the backup to the standby server
once backup is transferred, user resumes the CSD process
When choosing Transportable Media (1), parallel option and logshipping options are by default disabled:
Dbvisit will create backup in Source temp folder location and pause the CSD process. For resuming the CSD process, backup must be present in Target temp location on standby server.
All other options must be set and CSD process initiated as in normal CSD process which is described on this page. After the backup phase is finished, you will see the task status as follows:
The task details will then contain further instructions:
Once backup is placed into desired directory on standby server, you will be able to resume the task:
On the next page, confirm the resuming of the CSD: