Checkpoints: Synchronizing a Database and Its Log

A checkpoint writes all dirty pages—pages that have been modified in memory, but not on disk, since the last checkpoint—to the database device.

The SAP ASE automatic checkpoint mechanism guarantees that data pages changed by completed transactions are regularly written from the memory cache to the database device. Synchronizing the database and its transaction log shortens the time it takes to recover the database after a system failure.

• Setting the Recovery Interval
  Typically, automatic recovery takes anywhere between a few seconds and a few minutes per database.
• Automatic Checkpoint Procedure
  Approximately once a minute, a checkpoint task checks each database on the server to see how many records have been added to the transaction log since the last checkpoint.
• Truncating the Log After Automatic Checkpoints
  System administrators can truncate the transaction log when SAP ASE performs an automatic checkpoint.
• Free Checkpoints
  When SAP ASE has no user tasks to process, a housekeeper wash task automatically begins writing dirty buffers to disk. Checkpoints that occur as a result of the housekeeper wash task are known as free checkpoints.
• Manually Requesting a Checkpoint
  Database owners can issue the checkpoint command to force all modified pages that are in memory to be written to disk.