ABAP Keyword Documentation →  ABAP - Reference →  Processing External Data →  Data Consistency → 

Database Locks

Any database permitting simultaneous access to multiple transactions requires database locks to manage and synchronize access. The tasks of this mechanism are to:

Example

A booking needs to be made in a flight reservation system for Lufthansa flight 0400 on May 16, 1996. This is possible only if there are enough free seats. To prevent two bookings from being made at the same time and avoid overbooking, the entry in the database table SFLIGHT for this flight must be locked to prevent it from being changed by other transactions. This ensures that the query to determine the number of free seats in the SEATSOCC field can be carried out, the flight can be booked, and the SEATSOCC field can be updated by other transactions.

How does the lock concept work?

Database systems do not usually provide commands for explicitly setting or releasing locks. Database locks are therefore set implicitly when one of the Open SQL statements SELECT, INSERT, UPDATE, MODIFY, or DELETE (or the corresponding Native SQL statement) is called before the database operation is executed.

What is locked?

Database systems set physical locks: All rows affected by a database call are locked. In the case of SELECT, these are the selected entries. In the case of UPDATE, DELETE, INSERT, and MODIFY, these are the lines to be changed, deleted, or inserted.

The following call

SELECT SINGLE FOR UPDATE * FROM sflight
  WHERE
    CARRID   = 'LH'       AND
    CONNID   = '0400'     AND
    FLDATE   = '19960516'.

for example, locks the entry in table SFLIGHT for Lufthansa flight 0400 on May 16, 1996.

It is not always the table row that is locked. Tables, data pages, and index pages can also be locked, for example. The units locked depend on the database system you are using and the action being carried out.

Lock Mode

In principle, one type of lock is enough to control competing data access. However, to allow a larger number of transactions to run in parallel, database systems use a range of lock types. These can vary from system to system, but the following two examples outline sufficiently how locks work:

How are locks set?

Exclusive locks are set by the Open SQL statements SELECT SINGLE FOR UPDATE, INSERT, UPDATE, MODIFY, and DELETE (or the appropriate Native SQL statements).

Whether the Open SQL command SELECT (or the corresponding Native SQL command) sets a lock depends on the isolation level of the transaction. There are two possible isolation levels:

Many database systems employ additional isolation levels (such as "cursor stability" and "repeatable read"). These work in a similar way to "committed read", but the shared lock is retained until the next data object is read or until the database cursor is closed. Since these isolation levels are not sufficiently standardized, they are not currently used in AS ABAP.

If a transaction cannot lock an object because it is already locked by another transaction, it waits until the other transaction has released the lock. This can result in a deadlock. A deadlock occurs, for example, when two transactions are waiting for a lock held by the other.

The following program code demonstrates a solution to this problem:

DATA sflight_wa TYPE sflight, sbook_wa type sbook.

SELECT SINGLE FOR UPDATE * FROM sflight
       INTO sflight_wa
  WHERE
    carrid   = 'LH'       AND
    connid   = '0400'     AND
    fldate   = '19960516'.
IF sy-subrc <> 0.
  MESSAGE e...
ENDIF.

IF sflight_wa-seatsocc < sflight_wa-seatsmax.
  sbook_wa-carrid = 'LH'.
  sbook_wa-connid = '0400'.
  sbook_wa-fldate = '19960516'.
  ...
  INSERT sbook FROM sbook_wa.
  IF sy-subrc <> 0.
    MESSAGE e...
  ENDIF.
  UPDATE sflight
    SET
      seatsocc = seatsocc + 1
    WHERE
      carrid   = 'LH'       AND
      connid   = '0400'     AND
      fldate   = '19960516'.
ELSE.
  MESSAGE e...
ENDIF.

COMMIT WORK.

The table row selected by SELECT SINGLE FOR UPDATE and inserted by INSERT is locked until the end of the database LUW. This prevents the flight from being overbooked and inconsistencies from occurring between tables SFLIGHT and SBOOK in the event of a database rollback after an error.

How long is a lock retained?

All database locks are released no later than the next database commit or rollback. Shared locks are usually retained for a shorter period. Sometimes, this causes problems for transactions that involve multiple dialog steps:

After the user has selected a flight in the above example, he or she usually performs further dialog steps to enter additional data for the reservation. Here, the flight reservation is added in a different database LUW than the original selection of the flight. Database locking does not prevent another transaction from booking this flight in the meantime, which can mean that the scheduled booking may have to be canceled after all.

From the user's point of view, this solution is very inconvenient. To avoid this scenario, a flight reservation system must use the SAP locking mechanism to lock the flight for the entire duration of the transaction.



Continue
Lock Conflicts in IBM DB2