The Oracle database initialization parameters in the init.ora file are distinct from gateway initialization parameters. Set the gateway parameters in the initialization parameter file using an agent-specific mechanism, or set them in the Oracle data dictionary using the DBMS_HS package. The gateway initialization parameter file must be available when the gateway is started.
This appendix contains a list of the gateway initialization parameters that can be set for each gateway and their description. It also describes the initialization parameter file syntax. It includes the following sections:
The syntax for the initialization parameter file is as follows:
The file is a sequence of commands.
Each command should start on a separate line.
End of line is considered a command terminator (unless escaped with a backslash).
If there is a syntax error in an initialization parameter file, none of the settings take effect.
Set the parameter values as follows:
[SET][PRIVATE] parameter=value
Where:
parameter is an initialization parameter name. It is a string of characters starting with a letter and consisting of letters, digits and underscores. Initialization parameter names are case sensitive.
value is the initialization parameter value. It is case sensitive. An initialization parameter value is either:
A string of characters that does not contain any backslashes, white space or double quotation marks (")
A quoted string beginning with a double quotation mark and ending with a double quotation mark. The following can be used inside a quoted string:
backslash (\) is the escape character
\n inserts a new line
\t inserts a tab
\" inserts a double quotation mark
\\ inserts a backslash
A backslash at the end of the line continues the string on the next line. If a backslash precedes any other character then the backslash is ignored.
For example, to enable tracing for an agent, set the HS_FDS_TRACE_LEVEL initialization parameter as follows:
HS_FDS_TRACE_LEVEL=ON
SET and PRIVATE are optional keywords. You cannot use either as an initialization parameter name. Most parameters are needed only as initialization parameters, so you usually do not need to use the SET or PRIVATE keywords. If you do not specify either SET or PRIVATE, the parameter is used only as an initialization parameter for the agent.
SET specifies that, in addition to being used as an initialization parameter, the parameter value is set as an environment variable for the agent process. Use SET for parameter values that the drivers or non-Oracle system need as environment variables.
PRIVATE specifies that the initialization parameter should be private to the agent and should not be uploaded to the Oracle database. Most initialization parameters should not be private. If, however, you are storing sensitive information like a password in the initialization parameter file, then you may not want it uploaded to the server because the initialization parameters and values are not encrypted when uploaded. Making the initialization parameters private prevents the upload from happening and they do not appear in dynamic performance views. Use PRIVATE for the initialization parameters only if the parameter value includes sensitive information such as a user name or password.
SET PRIVATE specifies that the parameter value is set as an environment variable for the agent process and is also private (not transferred to the Oracle database, not appearing in dynamic performance views or graphical user interfaces).
This section lists all the initialization file parameters that can be set for the Oracle Database Gateway for SQL Server. They are as follows:
The following sections describe all the initialization file parameters that can be set for gateways.
| Property | Description |
|---|---|
| Default value | None |
| Range of values | Not applicable |
Specifies the remote functions that can be referenced in SQL statements. The value is a list of remote functions and their owners, separated by semicolons, in the following format:
owner_name.function_name
For example:
owner1.A1;owner2.A2;owner3.A3
If an owner name is not specified for a remote function, the default owner name becomes the user name used to connect to the remote database (specified when the Heterogeneous Services database link is created or taken from user session if not specified in the DB link).
The entries for the owner names and the function names are case sensitive.
| Property | Description |
|---|---|
| Default value | WORLD |
| Range of values | 1 to 199 characters |
Specifies a unique network sub-address for a non-Oracle system. The HS_DB_DOMAIN initialization parameter is similar to the DB_DOMAIN initialization parameter, described in the Oracle Database Reference. The HS_DB_DOMAIN initialization parameter is required if you use the Oracle Names server. The HS_DB_NAME and HS_DB_DOMAIN initialization parameters define the global name of the non-Oracle system.
Note:
TheHS_DB_NAME and HS_DB_DOMAIN initialization parameters must combine to form a unique address in a cooperative server environment.| Property | Description |
|---|---|
| Default value | 01010101 |
| Range of values | 1 to 16 hexadecimal characters |
Specifies a unique hexadecimal number identifying the instance to which the Heterogeneous Services agent is connected. This parameter's value is used as part of a transaction ID when global name services are activated. Specifying a nonunique number can cause problems when two-phase commit recovery actions are necessary for a transaction.
| Property | Description |
|---|---|
| Default value | HO |
| Range of values | 1 to 8 characters |
Specifies a unique alphanumeric name for the data store given to the non-Oracle system. This name identifies the non-Oracle system within the cooperative server environment. The HS_DB_NAME and HS_DB_DOMAIN initialization parameters define the global name of the non-Oracle system.
| Property | Description |
|---|---|
| Default value | 100 |
| Range of values | 1 to 4000 |
Specifies the maximum number of entries in the describe cache used by Heterogeneous Services. This limit is known as the describe cache high water mark. The cache contains descriptions of the mapped tables that Heterogeneous Services reuses so that it does not have to re-access the non-Oracle data store.
If you are accessing many mapped tables, increase the high water mark to improve performance. Increasing the high water mark improves performance at the cost of memory usage.
| Property | Description |
|---|---|
| Default value | System-specific |
| Range of values | Any valid language name (up to 255 characters) |
Provides Heterogeneous Services with character set, language, and territory information of the non-Oracle data source. The value must use the following format:
language[_territory.character_set]
Note:
The globalization support initialization parameters affect error messages, the data for the SQL Service, and parameters in distributed external procedures.Ideally, the character sets of the Oracle database and the non-Oracle data source are the same. In almost all cases, HS_LANGUAGE should be set exactly the same as Oracle database character set for optimal character set mapping and performance. If they are not the same, Heterogeneous Services attempts to translate the character set of the non-Oracle data source to the Oracle database character set, and back again. The translation can degrade performance. In some cases, Heterogeneous Services cannot translate a character from one character set to another.
Note:
The specified character set must be a superset of the operating system character set on the platform where the agent is installed.As more Oracle databases and non-Oracle databases use Unicode as database character sets, it is preferable to also run the gateway in Unicode character set. To do so, you must set HS_LANGUAGE=AL32UTF8. However, when the gateway runs on Windows, the Microsoft ODBC Driver Manager interface can exchange data only in the double-byte character set, UCS2. This results in extra ratio expansion of described buffer and column sizes. Refer to HS_FDS_REMOTE_DB_CHARSET for instruction on how to adjust to correct sizes.
The language component of the HS_LANGUAGE initialization parameter determines:
Day and month names of dates
AD, BC, PM, and AM symbols for date and time
Default sorting mechanism
Note that Oracle does not determine the language for error messages for the generic Heterogeneous Services messages (ORA-25000 through ORA-28000). These are controlled by the session settings in the Oracle database.
The territory clause specifies the conventions for day and week numbering, default date format, decimal character and group separator, and ISO and local currency symbols. Note that the level of globalization support between the Oracle database and the non-Oracle data source depends on how the gateway is implemented.
| Property | Description |
|---|---|
| Default value | 64 KB |
| Range of values | Any value up to 2 GB |
Sets the size of the piece of LONG data being transferred. A smaller piece size means less memory requirement, but more round-trips to fetch all the data. A larger piece size means fewer round-trips, but more of a memory requirement to store the intermediate pieces internally. Thus, the initialization parameter can be used to tune a system for the best performance, with the best trade-off between round-trips and memory requirements, and network latency or response time.
| Property | Description |
|---|---|
| Default value | 50 |
| Range of values | 1 to the value of Oracle's OPEN_CURSORS initialization parameter |
Defines the maximum number of cursors that can be open on one connection to a non-Oracle system instance.
The value never exceeds the number of open cursors in the Oracle database. Therefore, setting the same value as the OPEN_CURSORS initialization parameter in the Oracle database is recommended.
| Property | Description |
|---|---|
| Default value | ON |
| Range of values | OFF or ON |
Controls whether Heterogeneous Services attempts to optimize performance of data transfer between the Oracle database and the Heterogeneous Services agent connected to the non-Oracle data store.
The following values are possible:
OFF disables reblocking of fetched data so that data is immediately sent from agent to server.
ON enables reblocking, which means that data fetched from the non-Oracle system is buffered in the agent and is not sent to the Oracle database until the amount of fetched data is equal to or higher than the value of HS_RPC_FETCH_SIZE initialization parameter. However, any buffered data is returned immediately when a fetch indicates that no more data exists or when the non-Oracle system reports an error.
| Property | Description |
|---|---|
| Default value | 50000 |
| Range of values | 1 to 10000000 |
Tunes internal data buffering to optimize the data transfer rate between the server and the agent process.
Increasing the value can reduce the number of network round-trips needed to transfer a given amount of data, but also tends to increase data bandwidth and to reduce latency as measured between issuing a query and completion of all fetches for the query. Nevertheless, increasing the fetch size can increase latency for the initial fetch results of a query, because the first fetch results are not transmitted until additional data is available.
| Property | Description |
|---|---|
| Default value for '[+|-]hh:mm' | Derived from the NLS_TERRITORY initialization parameter |
| Range of values for '[+|-]hh:mm' | Any valid datetime format mask |
Specifies the default local time zone displacement for the current SQL session. The format mask, [+|-]hh:mm, is specified to indicate the hours and minutes before or after UTC (Coordinated Universal Time—formerly Greenwich Mean Time). For example:
HS_TIME_ZONE = [+ | -] hh:mm
| Property | Description |
|---|---|
| Default Value | COMMIT_CONFIRM |
| Range of Values | COMMIT_CONFIRM, READ_ONLY, SINGLE_SITE, READ_ONLY_AUTOCOMMIT, SINGLE_SITE_AUTOCOMMIT |
Specifies the type of transaction model that is used when the non-Oracle database is updated by a transaction.
The following values are possible:
COMMIT_CONFIRM provides read and write access to the non-Oracle database and allows the gateway to be part of a distributed update. To use the commit-confirm model, the following items must be created in the non-Oracle database:
Transaction log table. The default table name is HS_TRANSACTION_LOG. A different name can be set using the HS_FDS_TRANSACTION_LOG parameter. The transaction log table must be granted SELECT, DELETE, and INSERT privileges set to public.
Recovery account. The account name is assigned with the HS_FDS_RECOVERY_ACCOUNT parameter.
Recovery account password. The password is assigned with the HS_FDS_RECOVERY_PWD parameter.
READ_ONLY provides read access to the non-Oracle database.
SINGLE_SITE provides read and write access to the non-Oracle database. However, the gateway cannot participate in distributed updates.
READ_ONLY_AUTOCOMMIT provides read only access to the non-Oracle database that does not use logging.
SINGLE_SITE_AUTOCOMMIT provides read and write access to the non-Oracle database without logging. The gateway cannot participate in distributed updates. Moreover, any update to the non-Oracle database is committed immediately.
| Property | Description |
|---|---|
| Default value | None |
| Range of values | Valid parameter file names |
Use the IFILE initialization parameter to embed another initialization file within the current initialization file. The value should be an absolute path and should not contain environment variables. The three levels of nesting limit do not apply.
See Also:
Oracle Database Reference| Property | Description |
|---|---|
| Default Value | None |
| Range of Values | Not applicable |
HS_FDS_CONNECT_INFO that describes the connection to the non-Oracle system.
The default initialization parameter file already has an entry for this parameter. The syntax for HS_FDS_CONNECT_INFO for the gateway is as follows:
For UNIX:
HS_FDS_CONNECT_INFO=host_name[[:port_number]|/[instance_name]][/database_name]
where, host_name is the host name or IP address of the machine hosting the SQL Server database, port_number is the port number of the SQL Server, instance_name is the instance of SQL Server running on the machine, and database_name is the SQL Server database name.
Either of the variables port_number or instance_name can be used, but not both together. Optionally, they both can be omitted. The variable database_name is always optional. The slash (/) is required when a particular value is omitted. For example, all of the following entries are valid:
HS_FDS_CONNECT_INFO=host_name/instance_name/database_name HS_FDS_CONNECT_INFO=host_name//database_name HS_FDS_CONNECT_INFO=host_name:port_name//database_name HS_FDS_CONNECT_INFO=host_name/instance_name HS_FDS_CONNECT_INFO=host_name
For Windows:
HS_FDS_CONNECT_INFO= host_name/[instance_name][/database_name]
where, host_name is the host name or IP address of the machine hosting the SQL Server database, instance_name is the instance of SQL Server running on the machine, and database_name is the SQL Server database name.
Both instance_name and database_name are optional. If instance_name is omitted and database_name is provided, the slash (/) is required. This can be shown as follows:
HS_FDS_CONNECT_INFO= host_name//database_name
This release supports IPv6 format, so you can enter IPv6 format in place of hostname, but you need to wrap square brackets around the IPv6 specification.
For example,
HS_FDS_CONNECT_INFO=[2001:0db8:20c:f1ff:fec6:38af]:port_number/…
| Property | Description |
|---|---|
| Default Value | FALSE |
| Range of Values | TRUE, FALSE |
Enables return values from functions. By default, all stored procedures and functions do not return a return value to the user.
Note:
If you set this initialization parameter, you must change the syntax of the procedure execute statement for all existing stored procedures to handle return values.| Property | Description |
|---|---|
| Default Value | RECOVER |
| Range of values | Any valid user ID |
Specifies the name of the recovery account used for the commit-confirm transaction model. An account with user name and password must be set up at the non-Oracle system. For more information about the commit-confirm model, see the HS_TRANSACTION_MODEL parameter.
The name of the recovery account is case sensitive.
| Property | Description |
|---|---|
| Default Value | RECOVER |
| Range of values | Any valid password |
Specifies the password of the recovery account used for the commit-confirm transaction model set up at the non-Oracle system. For more information about the commit-confirm model, see the HS_TRANSACTION_MODEL parameter.
The name of the password of the recovery account is case sensitive.
| Property | Description |
|---|---|
| Default Value | FALSE |
| Range of Values | TRUE, FALSE |
Enables Oracle Database Gateway for SQL Server to treat SINGLE FLOAT PRECISION fields as DOUBLE FLOAT PPRECISION fields.
| Property | Description |
|---|---|
| Default Value | FALSE |
| Range of Values | TRUE, FALSE |
Enables result sets to be returned from stored procedures. By default, all stored procedures do not return a result set to the user.
Note:
If you set this initialization parameter, you must do the following:Change the syntax of the procedure execute statement for all existing stored procedures, to handle result sets
Work in the sequential mode of Heterogeneous Services
| Property | Description |
|---|---|
| Default Value | OFF |
| Range of values | OFF, ON, DEBUG |
Specifies whether error tracing is turned on or off for gateway connectivity.
The following values are valid:
OFF disables the tracing of error messages.
ON enables the tracing of error messages that occur when you encounter problems. The results are written by default to a gateway log file in LOG directory where the gateway is installed.
DEBUG enables the tracing of detailed error messages that can be used for debugging.
| Property | Description |
|---|---|
| Default Value | HS_TRANSACTION_LOG |
| Range of Values | Any valid table name |
Specifies the name of the table created in the non-Oracle system for logging transactions. For more information about the transaction model, see the HS_TRANSACTION_MODEL parameter.
| Property | Description |
|---|---|
| Default Value | 100 |
| Range of Values | Any integer between 1 and 1000 |
| Syntax | HS_FDS_FETCH_ROWS=num |
HS_FDS_FETCH_ROWS specifies the fetch array size. This is the number of rows to be fetched from the non-Oracle database and to return to Oracle database at one time. This parameter will be affected by the HS_RPC_FETCH_SIZE and HS_RPC_FETCH_REBLOCKING parameters.
| Property | Description |
|---|---|
| Default Value | 0 (no timeout) |
| Range of Values | 0-9999 (minutes) |
| Syntax | HS_IDLE_TIMEOUT=num |
This feature is only available for Oracle Net TCP protocol. When there is no activity for a connected gateway session for this specified time period, the gateway session would be terminated automatically with pending update (if any) rolled back.
| Property | Description |
|---|---|
| Default Value | BYTE |
| Range of Values | BYTE | CHAR |
| Syntax | HS_NLS_LENGTH_SEMANTICS = { BYTE | CHAR } |
This release of gateway has Character Semantics functionality equivalent to the Oracle Database Character Semantics, that is, NLS_LENGTH_SEMANTICS. When HS_NLS_LENGTH_SEMANTICS is set to CHAR, the (VAR)CHAR columns of SQL Server database are to be interpreted as having CHAR semantics. The only situation the gateway does not honor the HS_NLS_LENGTH_SEMANTICS=CHAR setting is when both Oracle database and the gateway are on the same multi-byte character set
| Property | Description |
|---|---|
| Default Value | OFF |
| Range of Values | OFF | LOCAL | REMOTE | ALL |
| Syntax | HS_KEEP_REMOTE_COLUMN_SIZE = OFF | LOCAL | REMOTE | ALL |
| Parameter type | String |
HS_KEEP_REMOTE_COLUMN_SIZE specifies whether to suppress ratio expansion when computing the length of (VAR)CHAR datatypes during data conversion from the non-Oracle database to the gateway, and then to the Oracle database. When it is set to REMOTE, the expansion is suppressed between the non-Oracle database and the gateway. When it is set to LOCAL, the expansion is suppressed between the gateway and the Oracle database. When it is set to ALL, the expansion is suppressed from the non-Oracle database to the Oracle database.
When the parameter is set, the expansion is suppressed when reporting the remote column size, calculating the implicit resulting buffer size, and instantiating in the local Oracle database. This has effect only for remote column size from non-Oracle database to Oracle database. If the gateway runs on Windows and HS_LANGUAGE=AL32UTF8, then you must not specify this parameter, as it would influence other ratio related parameter operation. It has no effect for calculating ratio for data moving from Oracle database to non-Oracle database through gateway during INSERT, UPDATE, or DELETE.
| Property | Description |
|---|---|
| Default Value | None |
| Range of values | Not applicable |
| Syntax | HS_FDS_REMOTE_DB_CHARSET |
This parameter is valid only when HS_LANGUAGE is set to AL32UTF8 and the gateway runs on Windows. As more Oracle databases and non-Oracle databases use Unicode as database character sets, it is preferable to also run the gateway in Unicode character set. To do so, you must set HS_LANGUAGE=AL32UTF8. However, when the gateway runs on Windows, the Microsoft ODBC Driver Manager interface can exchange data only in the double-byte character set, UCS2. This results in extra ratio expansion of described buffer and column sizes. To compensate, the gateway can re-adjust the column size if HS_FDS_REMOTE_DB_CHARSET is set to the corresponding non-Oracle database character set. For example, HS_FDS_REMOTE_DB_CHARSET=KO16KSC5601.
| Property | Description |
|---|---|
| Default Value | TRUE |
| Range of values | {TRUE|FALSE} |
| Syntax | HS_FDS_SUPPORT_STATISTICS= {TRUE|FALSE} |
We gather statistics from the non-Oracle database by default. You can choose to disable the gathering of remote database statistics by setting the HS_FDS_SUPPORT_STATISTICS parameter to FALSE.
| Property | Description |
|---|---|
| Default Value | FALSE |
| Range of values | {TRUE|FALSE} |
| Syntax | HS_FDS_RSET_RETURN_ROWCOUNT= {TRUE|FALSE} |
When set to TRUE, the gateway returns the row counts of DML statements that are executed inside a stored procedure. The row count is returned as a single row, single column result set of type signed integer.
When set to FALSE, the gateway skips the row counts of DML statements that are executed inside a stored procedure. This is the default behavior, and it is the behavior of 11.1 and older gateways.
| Property | Description |
|---|---|
| Default Value | 64 |
| Range of values | {64|32} |
| Syntax | HS_FDS_SQLLEN_INTERPRETATION= {64|32} |
This parameter is only valid for 64 bit platforms. ODBC standard specifies SQLLEN (of internal ODBC construct) being 64 bit on 64 bit platforms, but some ODBC driver managers and drivers violate this convention, and implement it as 32 bit. In order for the gateway to compensate their behavior, you need to specify HS_FDS_SQLLEN_INTERPRETATION=32 if you use these types of driver managers and driver.