Shadow tables for primary tables with large objects

When you create a primary article for a table with one or more large-object columns, ASE Replicator creates a shadow table with the columns described in Table A-18.

Table A-18: Shadow table with large object

Column name

Datatype

Description

conn_id

numeric(18,0)

ID of the primary connection

txid_page

int

Identifies the transaction ID page in the transaction log table

txid_row

smallint

Identifies the transaction ID row in the transaction log table

log_ts_high

smallint

Log timestamp high value in the transaction log table

log_ts_low

int

Log timestamp low value in the transaction log table

opid_page

int

Identifies the operation ID page in the transaction log table

opid_row

smallint

Identifies the operation ID row in the transaction log table

op_code

tinyint

Identifies type of operation

op_xstat

int

Identifies operation as either update or delete

image_type_

char(1)

Identifies operation type or procedure execution

pkey

numeric(5,0)

Primary key column(s) value

col1

Specified in primary object

Column for a published field in the primary object

col2

rpl_clob

Column for a published large-object (text) field in the primary object

...

coln

...

...

Additional column for each published field in the primary object

Except for large-object primary table columns, the shadow table column for each published field has the datatype of the published field, and each row contains the data from that primary column for the transaction operation identified by that row.

For each large-object primary table column, the shadow table column has a user-defined Boolean datatype that identifies the type of large-object data in the primary column, and the value of that Boolean indicates whether a change was made in the primary column data by the transaction operation identified by that row.

The user-defined Boolean datatypes are:

All these datatypes map to the Adaptive Server bit datatype.