Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Variable Name

Variable Description

Possible Values

Notes

WAIT_FOR_FILE_SEC

This variable defines timeout in seconds which dbvctl has for response for submitting SQL or RMAN command in primary or standby database. This includes tasks such as:

  • applying archivelog file with sqlplus on standby database

  • performing logswitch

  • logon with RMAN or sqlplus

etc …

Example:

WAIT_FOR_FILE_SEC = 250

125 (Default)

Edit with care

Info

Chainging WAIT_FOR_FILE_SEC is especially useful when applying of archivelogs takes too long on your standby database and you receive frequent errors. The symptom of which is the error message:

Code Block
file /xxxx/33593.dbvisit.201904160305.sqlplus.dbv) does not exist or is empty. Please check space and file permissions. 

In such case we recommend to use either APPLY_ARCHIVE_RMAN = Y or increase the timeout to 250 seconds. You always need to consider actual archivelog size - for example getting this tiemout for an archivelog of 100MB size (meaning apply took more than this timeout) means that you have performance issue on your standby server.

RMAN_DBF_BACKUP_TYPE

The Create Standby Database process (CSD) is by default doing RMAN COMPRESSED BACKUPSET which favors environments with slow networks and low free disk space capacity.

As result, compressed backup takes more time to be created and restored.

If your network speed and free disk space are sufficient, we recommend to switch to “AS BACKUPSET” or “AS COPY” to decrease the time needed to create standby database.

Windows platform has very poor performance for RMAN COMPRESSED BACKUPSET.

You can always set the parameter before standby database creation. You may need to create standby database several times before finding optimal setting.

AS BACKUPSET

RMAN backups done without compression
AS COMPRESSED BACKUPSET (Default)

RMAN backups are done as compressed backupset
AS COPY

Direct Datafile copy is used for standby creation

APPLY_READ_ONLY

Specifies behavior for situation when archivelog apply job is run and standby database is found in READ ONLY mode.

By default disabled, meaning standby database will stay in READ ONLY mode and apply job will fail.

Enablig this variable (setting to Y) will forcefully restart standby database to MOUNT (recovery mode) if it is found in READ ONLY state during archivelog apply.

Standby database will remain in MOUNT mode after archivelog is applied

N (Default)

Standby database will remain in READ ONLY mode

Y

Force restart to MOUNT mode during apply

Info

setting APPLY_READ_ONLY = Y is especially useful on Windows systems, where after server reboot the standby database will be typically started in READ ONLY mode. Without this paramerter set, archivelog apply will be halted (fail indefinitely) and your standby database will get out of sync.

LOGDIR

Directory on primary host in which Dbvisit generates various small logfiles

[DBVISIT_BASE]/standbymp/oracle/log (Default)

LOGDIR_DR

Directory on standby host in which Dbvisit generates various small logfiles

[DBVISIT_BASE]/standbymp/oracle/log (Default)

TRCDIR

Directory on primary host in which Dbvisit generates dbvctl traces

[DBVISIT_BASE]/standbymp/oracle/trace (Default)

TRCDIR_DR

Directory on standby host in which Dbvisit generates dbvctl traces

[DBVISIT_BASE]/standbymp/oracle/trace (Default)

Info

If you require to change these 4 variables (for example due to lack of disk space on [DBVISIT_BASE] mountpoint, It makes sense to point all 4 variables LOGDIR, LOGDIR_DR, TRCDIR and TRCDIR_DR into same parent directory

RUN_INSPECT

By default, each time DMN_DBVISIT transfer archivelog job runs, log gap report is run as well. Log gap report always results in several network roundtrips and in environments with slow response networks, this can significantly prolong archivelog transfer.

Set this value to N if you wish to skip the log gap report during transfer. This will affect ARCHIVE_LOG_GAP_THRESHOLD and TRANSFER_LOG_GAP_THRESHOLD notifications.

N

No log gap report is run before transfer

Y (Default)

Log gap report is run before transfer.

Edit with care

SYNCH_DBVISIT_INSTALL

This variable ensures the default behavior that DDC file on standby is overwritten with DDC file from primary whenever differences occur.

Setting this variable to N will cause the standby DDC file to never be overwritten, which can cause potential issues.

This value is for example utilized for DR Test functionality on standby. You shouldn’t change this variable unless instructed by Dbvisit Support

N

DDC file on standby will not be overwritten

Y (Default)

DDC file on standby will be overwritten by DDC from primary

Not supported to edit

DB_CONNECTION_MODE

Determines the type of SQLPLUS and RMAN connections Dbvisit Standby makes to the database on the standby server.

By default in shared mode, the connection is persistent and will be re-used by Dbvisit tasks.

Dedicated connection mode will force the disconnection and creation of new session each time Dbvisit will need to run sqlplus or RMAN session.

shared (default)

dedicated

DB_CONNECTION_MODE_DR

same as DB_CONNECTION_MODE, but for standby database

shared (default)

dedicated

LOG_SWITCH_WAIT_IN_SEC

Wait in seconds to allow log switch to occur. This timeout is defined from the moment the logswitch command is issued, till the completion (archivelog visible in v$archived_log).

3 (default)

RETRIES_WAIT_IN_SEC

Time in seconds to wait before performing another transfer if a previous transfer failed.

5 (default)

GSDIR

Directory on Primary server to transfer online redo and other tasks during Graceful switchover.

default location is in [DBVISIT_BASE]. You should change this variable, if you online redo member size exceeds available free disk space in [DBVISIT_BASE] mountpoint

[DBVISIT_BASE]/standbymp/oracle/gs (Default)

Edit with care

GSDIR_DR

Directory on Standby server to transfer online redo and other tasks during Graceful switchover.

default location is in [DBVISIT_BASE]. You should change this variable, if you online redo member size exceeds available free disk space in [DBVISIT_BASE] mountpoint

[DBVISIT_BASE]/standbymp/oracle/gs (Default)

Edit with care

RECOVER_UNTIL_SCN_FILE

This variable defines file which (if exists), will halt the standby database recovery at the specified SCN. By default, the file specified by default value doesn’t exist.

Example:

Code Block
$ cat /usr/dbvisit/standbymp/oracle/log/recover_until_scn.txt
221862819

any archivelog apply job on standby database will then only recover standby database to specified time.

Apply job will output:

Code Block
>>> Applying Log file(s) from czlin0231 to SLASH on czlin0232:

    Taking into account RECOVER_UNTIL_SCN=221862819 set in file
    /usr/dbvisit/standbymp/oracle/log/recover_until_scn.txt

[DBVISIT_BASE]/standbymp/oracle/log/recover_until_scn.txt (Default)

Edit with care

RECOVER_UNTIL_TIME_FILE

This variable defines file which (if exists), will halt the standby database recovery at the specified time in format YYYY-MM-DD:HH24:MI:SS

By default, the file specified by default value doesn’t exist.

Example:

When following file is created:

Code Block
$ cat /usr/dbvisit/standbymp/oracle/log/recover_until_time.txt
2025-02-21:11:43:00

any archivelog apply job on standby database will then only recover standby database to specified time.

Apply job will output:

Code Block
>>> Applying Log file(s) from czlin0231 to SLASH on czlin0232:

    Taking into account RECOVER_UNTIL_TIME=2025-02-21:11:43:00 set in file
    /usr/dbvisit/standbymp/oracle/log/recover_until_time.txt

[DBVISIT_BASE]/standbymp/oracle/log/recover_until_time.txt (Default)

Edit with care

Note

The Standby database will never be recovered past the time or SCN specified in RECOVER_UNTIL_SCN_FILE and in RECOVER_UNTIL_TIME_FILE until those files cease to exist or until they are updated.

These two variables affect ASU apply jobs as well as manual apply jobs equally.

SEND_LOG_GAP_WITH_HEARTBEAT

Include the Dbvisit Standby log gap report in the daily heartbeat email that is sent from the primary server.

N

Y (Default)

SERVICE_USED

If your standby or primary database are registered with srvctl and has custom service, the value of this parameter has to be set to Y

This will result in adding a -f/force flag to the srvctl "stop instance" command to avoid below error.

Code Block
CRS-2529: Unable to act on 'ora.cpms.db' because that would require stopping or
relocating 'ora.cpms.cpmsdb_srv.svc', but the force option was not specified

N (Default)

Y

SERVICE_USED_DR

set to Y if SERVICE_USED is set to Y

N (Default)

Y

LOCK_MAX_TIME_SEC

If Dbvisit Standby detects a lock on a file that it is trying to process, this parameter specifies the maximum wait in seconds Dbvisit Standby will retry the lock.

If the file is still locked after LOCK_MAX_TIME_SEC, then Dbvisit Standby will raise an alert.

10 (Default)

LOCK_RETRY_TIMEOUT_MILLISEC

If Dbvisit Standby detects a lock on a file that it is trying to process, this parameter specifies the wait in milliseconds before Dbvisit Standby will attempt to access the file again. On some platforms the minimum timeout time may be 1 second.

500 (Default)

Info

Parameters LOCK_MAX_TIME_SEC and LOCK_RETRY_TIMEOUT_MILLISEC can be useful if your archivelog RMAN backups on primary interfere with the Dbvisit ASU DMN_DBVISIT transfer jobs. This usually manifests as error:

Code Block
File /xxxx/1_26428_1144409189.arc is in use or locked by another process. 

If this is your case, we recommend to increase the parameter LOCK_MAX_TIME_SEC to 60 seconds.