Please enable JavaScript.

Coggle requires JavaScript to display documents.

Recovery System - Coggle Diagram

- - - - Transaction identifier
      - Data-item identifier
      - Old value
      - New value
  - - - search process is time-consuming.
      - Most of the transactions that, according to our algorithm, need to be redone have already written their updates into the database.
- - - - The transaction can no longer proceed normally execution due to some internal condition, such as bad input, data not found, overflow or resource limit exceeded.
    - - The system has entered an undesired state, so a transaction cannot continue its normal execution.
  - - - hardware failures and software failures halt the system, but do not damage non-volatile storage contents
- - - - Logging of information about operations is called
    - - corresponding log
        records for physical logging
    - - logging of old-value and new-value information
  - - - Physical log records found during the scan are handled as described above, except those that are skipped as described shortly.
      - Completed logical operations, identified by operation-endrecords, are rolled back differently.
      - If the system finds a <Ti, Oj, cancel operation> record, it skips all previous records until it finds the <Ti, Oj, start operation> record.
      - As before, when the log record <Ti start> is found, the transaction rollback is complete and the system adds a record of <Ti abort> to the log.
- - - - Before a block of data in main memory can be committed to the database, all records related to the data in that block must have been committed to stable storage.
  - - - transactions would force-output all modified blocks to
        disk when they commit
    - - allows a transaction to commit even if it has modified some
        blocks that have not yet been written back to disk.
    - - blocks modified by a transaction that is still
        active should not be written to disk.
    - - allows the system to write changed blocks to disk even if not all the transactions that made those modifications have been committed.
    - - locks that are held for a short duration
    - - blocks that have been modified in the buffer but have not
        been subsequently output
  - - - The database system reserves part of main memory to serve as a buffer that it, rather than the operating system, manages.
      - The database system implements its buffer within virtual memory provided by the operating system.
    - - The operating system reserves disk space to store virtual memory pages that are not currently in main memory.
- - - - ARIES allows the logging of operations that should not
        be undone even if a transaction gets rolled back
    - - Some pages can be retrieved regardless of others, so that they can be used even while other pages are being retrieved.
    - - Transactions can record savepoints, and can be rolled back partially, up to a savepoint.
    - - The ARIES recovery algorithm can be used with index
        concurrency-control algorithms
    - - Redo can be postponed on a page being fetched from disk, and performed when the page is fetched.
- - - - When the primary fails, the backup site takes over processing and becomes the new primary.
    - - If the log at the remote backup grows large, recovery will
        take a long time.
    - - is important for the remote backup system to detect
        when the primary has failed.
    - - One-safe
        
        A transaction commits as soon as its commit log record is written to stable storage at the primary site.
      - Two-very-safe
        
        A transaction commits as soon as its commit log record is
        written to stable storage at the primary and the backup site.
      - Two-safe.
        
        This scheme is the same as two-very-safe if both primary and
        backup sites are active.
- - - - Redo Phase
        
        the system reproduces updates of all transactions by scanning the log from the last checkpoint.
        
        The logs that are replayed include transaction logs that were rolled back before the system crash and those that were not committed when the system crash occurred.
      - Undo phase
        
        the system rolls back all transactions in the undo list. It performs the rollback by scanning the registry backwards from the end.
        
        repeating history
        
        The actions are repeated in the same order in which they were originally performed.
- - - - Successful completion
      - Partial failure.
      - Total failure
  - - - Physical blocks, blocks residing on the disk
      - Buffer blocks, blocks residing temporarily in main memory
      - disk buffer, area of memory where blocks reside temporarily