Checkpoints are used to reduce the amount of work that needs to be performed during restart recovery. Checkpoints can be taken asynchronously ( i.e while updates are going on ). Such a fuzzy checkpoint is initiated by writing a begin-checkpoint record. Then the end-chkpt record is constructed by including in it the contents of the normal transaction table , the BP dirty-pages table and file mapping information for the objects. This record is then written to the log. Once that record reaches stable storage, the LSN of the begin-chkpt record is stored in the master record which is in a well known place on stable storage. ARIES does not require that any dirty pages be forced to nonvolatile storage during a checkpoint. The assumption is that the buffer manager is on a continuous basis , writing out dirty pages in the background using system processes.