"H" to "J" Configuration Settings

If your older mainframe system uses too many CPU cycles running DataSync, contact Technical Support and, if appropriate, we will help you change your hashing routine.

To enable the selection of the hash function for incremental updates, enter one of the following values:

If performing data synchronization using CNXHASH and HASH, the hash algorithm used by both client and server must match, or a full synchronization is performed for every sync operation.

For Adabas, the DBID of the database normally overrides the DBID specified at the table level. If HONORDBIDFILEID = 1, CONNX uses the DBID specified in the CDD at the table level.

Default - HONORDBIDFILEID = 0

This setting allows you to choose the compression and decompression method for PowerFlex applications that use DataFlex tables.

Normally, CONNX uses DataFlex compression and decompression methods for DataFlex data. PowerFlex has a different compression method than the one used by DataFlex. If this setting is non-zero, CONNX will use the PowerFlex method to compress and decompress DataFlex tables.

Use HuffmanPowerFlex only if your DataFlex compressed files were designed using PowerFlex.

Default - HUFFMANPOWERFLEX = 0 (use DataFlex methods to perform compression and decompression)

For Adabas, if this is 1, ET commands will be suppressed even if explicitly requested by the client unless insert, update, or delete was performed.

Default - IGNORECOMMITONREAD = 1

If an Adabas file contains MUs or PEs, the MUs and PEs are normalized into separate tables during the FDT import process. In addition, CONNX creates a table with "_FLAT" appended to the name. This "flat" table contains fields for the MUs and PEs within each record. If you don't want the normalized tables imported, you can set this setting to 1.

Default - IMPORTFLATONLY = 0 (Import all tables, not just the flattened tables)

If set to 1, CONNX will import the group names.

Default = IMPORTGROUPNAMES = 0

This setting is used to determine whether to replace the existing CDD definitions.

Default - IMPORTOVERWRITE = 1

This settings allows you to add to the prefixes that are attached to COBOL FD imported fields. Usage:

Use this setting in conjunction with the REPLACEPREFIX setting to exchange the REPLACEPREFIX string with the IMPORTPREFIX string.

If REPLACEPREFIX does not exist (or is ""), IMPORTPREFIX prefaces the field name. For example:

If the field is "BBB_Customer" and:

the field in the CDD would be "AAA_BBB_Customer".

During the CONNX Data Dictionary import process, all tables, views, and stored procedures can be imported into the CDD. Set this setting to 0 if you want to remove all the stored procedures from the Import Table Selection window.

Default - IMPORTPROCEDURES = 1 (Display Stored Procedures in Import Table Selection window)

During the CONNX Data Dictionary import process, all tables, views, and stored procedures can be imported into the CDD. Set this setting to 0 if you want to remove all the stored views from the Import Table Selection window.

Default - IMPORTVIEWS = 1 (Display Import Views in Import Table Selection window)

This Unix-only language setting contains the location of the Informix run-time libraries.

Please consult your Informix documentation for information regarding specific valid values for your operating system.

This setting is used by the CONNX Installation utility and does not need to be edited. It specifies whether INI settings from windows 95 and below have been converted to the windows registry.

This setting is used by the CONNX Installation utility and does not need to be edited. The location of the CONNX installation. When looking at the 32 bit registry it is the 32 bit CONNX file location. When looking at the 64 bit registry it is the 64 bit CONNX file location.

This Windows-only setting allows an administrator to pre-create Windows domain based User IDs in a CDD. This is useful in environments where the administrator wishes to use maximum CDD security. When using NT Integrated security in the default mode, the first time a user connects to a CDD, a special unique NT_INTEGRATED User ID will be created for them in the CDD. The NT_INTEGRATED User ID is based on their Windows credentials. If you wish to lock the CDD, so no User ID's can be created, do the following

To use this feature, :

If set to 1, this setting enables total connection pooling (including pooling of the CDD) at the JDBC level.

Default - JDBCCONNECTIONPOOLING = 0

If JDBCCONNECTIONPOOLING is enabled, this setting sets the pooling timeout in milliseconds.

Default - JDBCCONNECTIONPOOLINGTIMEOUT = 300000 (30 seconds)

This setting is used by the CONNX Installation utility and does not need to be edited. It specifies whether the JDBC Server Service was selected during the last installation.

This setting determines how many memory map views are created. The total number of memory map views is log(# of rows) * JOINCACHEMULTIPLIER.

Default - JOINCACHEMULTIPLIER = 3

You can reduce this amount to increase performance.

This setting determines how much memory is used to store key data when performing in-memory binary searches.

Default - JOINCACHESIZE = 16000000 (16 million)

This setting determines how many keys to use in a single pass when performing joins. When using Rdb 6.x, the recommended setting is 8. The maximum amount is 250. Usage:

Default - JOINCOUNT = 250

This setting determines if descriptors on MU and PE tables will be used in join optimization.

Here is an example - a where clause of "value in ('A00', 'A10')" - the value field is a PE with a descriptor on it. The first retrieval for 'A00' may bring back ISN 1 2 8 The second retrieval for 'A10' may bring back ISN 3 8 10

ISN #8 has 2 occurrences, one with value 'A00' and one with value 'A10'. This causes incorrect SQL records when the same ISN is returned multiple times for different values.

If it is known for certain that there is a unique set of values in the MU across all ISNs, this setting can safely be set to 1.

Default - JOINMUDESC = 0

0 = fnv hash routine	If CNXHASH is set to 0, or the configuration setting for HASH is 0, CONNX uses the Fowler / Noll / Vo (FNV) Hash algorithm.
1 = oy_djb 64bit hash routine	If CNXHASH is set to 1, or the configuration setting for HASH is 1, CONNX uses the Ozan Yigit (SDBM) Hash combined with the Daniel J. Bernstein Hash algorithm.
2 = bj 64bit hash routine	If CNXHASH is set to 2, or the configuration setting for HASH is 2, CONNX uses the Bob Jenkins Hash algorithm.
3 = oy 32bit hash routine	If CNXHASH is set to 3, or the configuration setting for HASH is 3, CONNX uses the Ozan Yigit (SDBM) Hash algorithm.
4 = djb 32bit hash routine	If CNXHASH is set to 4, or the configuration setting for HASH is 4, CONNX uses the Daniel J. Bernstein Hash algorithm.
5 = returns zero	If CNXHASH is set to 5, or the configuration setting for HASH is 5, CONNX uses the NULL hash algorithm. The NULL hash does nothing and returns zero. It is only for benchmarking purposes and cannot be used for any other purpose.
6 = oy_djb 32bit + djb 32bit hash routine	If CNXHASH is set to 6, or the configuration setting for HASH is 6, CONNX uses an alternate form of the Ozan Yigit (SDBM) Hash combined with the Daniel J. Bernstein Hash algorithm.
7 = u_mac 64bit hash routine	If CNXHASH is set to 7, or the configuration setting for HASH is 7, CONNX uses the UMAC Hash algorithm if the server is NOT a VAX. If the server is a VAX, the program will fire an assert() and stop if data synchronizations are attempted.

0	Existing table will not be overwritten when re-importing.
1	Overwrite existing tables when re-importing.
2	Preserve Column Long Names.

0	Default. System generated unique User ID.
1	Reserved for internal use.
2	Administrator generated User ID - must follow the format described in the text above.