Recovery system in transaction management

1. RELATIONAL DATABASE MANAGEMENT SYSTEM
TRANSACTION MANAGEMENT
By,
S.Subhalakshmi,
M.Sc(cs).
Nadar saraswathi college of arts and science.

2. TRANSACTION MANAGEMENT:
 A Transaction is a very small unit of a program and it
may contain several low level tasks.
 A Transaction in a database system must maintain
Atomicity, Consistency, Isolation, and Durability − commonly
known as ACID properties.
 In order to ensure accuracy, completeness, and data
integrity.

3. FAILURE CLASSIFICATION:
 The simplest type of failure is one that does not result
in the loss of information in the system.
 There are various types of failure that may occur in a
system.
 The failures that are more difficult to deal with are
those that result in loss of information.
 The following types of failures are:

4. TRANSACTION FAILURE:
 There are two types of errors that may cause a transaction to
fail:
 LOGICAL ERROR :transaction cannot complete due to
some internal error condition.
 SYSTEM ERROR: the database system must terminate an
active transaction due to an error condition (e.g., deadlock).

5.  SYSTEM CRASH: a power failure or other hardware or
software failure causes the system to crash.
 FAIL-STOPASSUMPTION: non-volatile storage contents
are assumed to not be corrupted by system crash.
 DISK FAILURE: a head crash or similar disk failure destroys
all or part of disk storage.
 DESTRUCTION IS ASSUMED TO BE
DETECTABLE: disk drives use checksums to detect failures.

6. RECOVERY WITH CONCURRENT
TRANSACTIONS :
 We modify the log-based recovery schemes to allow
multiple transactions to execute concurrently.
 All transactions share a single disk buffer and a single
log.
 A buffer block can have data items updated by one or
more transactions.

7. CHECKPOINTS:
 We modify the log-based recovery schemes to allow multiple
transactions to execute concurrently.
 All transactions share a single disk buffer and a single log
 A buffer block can have data items updated by one or more
transactions.
 Streamline recovery procedure by periodically performing
check pointing.

8. RESTART RECOVERY:
 When the system recovers from a crash, it constructs two list:
 The undo-list consists of transaction to be done.
 The system constructs the two lists as follows:
 Initially, they are both empty.
 The system scans the log backward, examining each record, until
it finds the first <checkpoint> record:

9.  For each record found of the form<Ti commit>,it adds Ti to redo-
list.
 For each record found of the form<Ti start>,if Ti is not in redo-
list, then it adds Ti to undo-list.
 When the system has examined all the appropriate log records, it
checks the list L in the checkpoint record.
 For each transaction Ti in L, if Ti is not in redo-list, then it adds
Ti to undo-list.

10. SHADOW PAGING:
 Shadow paging is an alternative to log-based recovery.
 IDEA: maintain two page tables during the lifetime of a
transaction .
 Store the shadow page table in nonvolatile storage, such that state
of the database prior to transaction execution may be recovered.
 Only current page table is used for data item accesses during
execution of the transaction.

12.  TO COMMIT A TRANSACTION :
1. Flush all modified pages in main memory to disk
2. Output current page table to disk
3. Make the current page table the new shadow page table, as
follows:
 keep a pointer to the shadow page table at a fixed (known)
location on disk.
 to make the current page table the new shadow page table,
simply update the pointer to point to current page table on disk.

13.  BUFFER MANAGEMENT:
 LOG RECORD BUFFERING: log records are buffered in
main memory, instead of being output directly to stable storage.
 Log force is performed to commit a transaction by forcing all its
log records (including the commit record) to stable storage.
 Several log records can thus be output using a single output
operation, reducing the I/O cost.

14.  DATABASE BUFFERING :
 Database maintains an in-memory buffer of data blocks.
 When a new block is needed, if buffer is full an existing block
needs to be removed from buffer.
 If the block chosen for removal has been updated, it must be
output to disk.
 If a block with uncommitted updates is output to disk, log
records with undo information for the updates are output to the
log on stable storage first.

15.  OPERATING SYSTEM ROLE IN BUFFER
MANAGEMENT:
 We can manage the database buffer by using one of this
approaches:
 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 the virtual
memory provided by the operating system.
 The database system in turn force-output the buffer blocks to the
database, after writing relevant log records to stable storage.

16.  Unfortunately, almost all current-generation operating systems
retain complete control of virtual memory.
 The operating system reserves space on disk for storing virtual-
memory pages that are not currently in main memory ; this space
is called swap space.
 Although both approaches suffer from some drawbacks, one or
the other must be chosen unless the operating system is designed
to support the requirements of database logging.