commit

Ends a transaction, preserving changes made to the database during the transaction.

exec sql [at connection_name]   
commit [transaction | tran | work] 
[transaction_name];

transaction | trans | work: The keywords transaction, trans, and work are interchangeable in the rollback statement, except that only work is ANSI-compliant.

/*

 ** Using chained transaction mode,

 ** synchronize tables on two servers

*/

 exec sql begin declare section;

     char     title_id[7];

     int      num_sold;

 exec sql end declare section;

     long     SQLCODE;

...

 try_update:

 exec sql whenever sqlerror goto abort_tran;

 exec sql at connect1 select sum(qty)

     into :num_sold

     from salesdetail

     where title_id = :title_id;

 exec sql at connect2 update current_sales

     set num_sold = :num_sold

     where title_id = :title_id;

exec sql at connect2 commit work;

 exec sql at connect1 commit work;

 return;

abort_tran:

 printf("oops, should have used 2-phase commit\n");

 exec sql whenever sqlerror continue;

 exec sql at connect2 rollback work;

 exec sql at connect1 rollback work;

 goto try_update;

This reference page mainly describes aspects of the Transact-SQL commit statement that differ when used with Embedded SQL. See the Adaptive Server Enterprise Reference Manual for more information about commit and Transact-SQL transaction management.
Transaction names must conform to the Transact-SQL rules for identifiers. Transaction names are a Transact-SQL extension: they cannot be used with the ANSI-compliant keyword work.
When nesting transactions, assign a transaction name only to the outermost begin transaction statement and its corresponding commit transaction or rollback transaction statement.