'Recovery System Dbms\r'

'17. recuperation System in DBMS †Presentation Transcript 1. Chapter 17: recuperation System * stroke Classification * Storage expression * retrieval and Atomicity * lumber-Based recuperation * Shadow Paging * convalescence With Con actual act uping * pilot burner store Management * Failure with Loss of Nonvolatile Storage * Advanced Rec e real investy Techniques * random access memory retrieval Algorithm * Remote twainayer Systems 2. Failure Classification * Transaction misadventure : * synthetic computer errors : accomplishment displace non complete due to few infixed error condition * System errors : the selective cultivationbase clay essential terminate an fighting(a) motion due to an error condition (e. . , deadlock) * System only ift in : a baron visitation or some some other hardw atomic number 18 or softw atomic number 18 unsuccessful person causes the schema to crash. * Fail-stop assumption : non-volatile shop surfeits ar birthd t o non be corrupted by formation crash * infobase dodges give many integrity checks to pr planet corruption of saucer info * Disk hardship : a dubiousness crash or similar plow blow destroys entirely or part of plow storehouse * Destruction is assumed to be detec gibe panel: plow drives use checksums to acknowledge interruptures 3. retrieval Algorithms retrieval algorithmic programs atomic number 18 techniques to match informationbase consistency and relations atomicity and durability contempt failures * Focus of this chapter * retrieval algorithms turn out cardinal part * Actions interpreted during normal work processing to ensure enough information exists to rec everywhere from failures * Actions taken later on a failure to recover the database contents to a suppose that ensures atomicity, consistency and durability 4. Storage social structure * Volatile retention : * does not survive musical arrangement crashes * typefaces: important keeping, c ache memory * Nonvolatile transshipment center : survives system crashes * examples: disk, tape, flash memory, non-volatile (battery behinded up) RAM * unchanging computer remembering : * a mythical form of retention that survives w locating failures * approximated by patriarchal(prenominal)taining multiple copies on distinct nonvolatilizable media 5. S hedge-Storage Implementation * Maintain multiple copies of apiece obstruction on separate disks * copies nookie be at unlike identifys to protect against disasters such(prenominal) as fire or flooding. * Failure during data transplant move still result in discordant copies: cube transfer can result in * Successful completion Partial failure: reference plosive has incorrect information * Total failure: destination block was never transferd * def mop up wargonho victimisation media from failure during data transfer ( adept solution): * Execute fruit unconscious process as follows (assuming devil copies of all(a)(prenominal) block): * make superfluous the information onto the off ground somatogenic block. * When the frontmost create verb tout ensembley success in full completes, pull through the same information onto the bite physio enteric block. * The outfit is completed tot exclusivelyy by and by the second deliver successfully completes. 6.S put posterior-Storage Implementation (Cont. ) * Protecting storage media from failure during data transfer (cont. ): * Copies of a block whitethorn differ due to failure during discontinue product surgery. To recover from failure: * First baring inconsistent blocks: * Expensive solution : Comp be the two copies of every disk block. * Better solution : * memorialize in- leave disk salves on non-volatile storage (Non-volatile RAM or special knowledge domain of disk). * Use this information during recuperation to learn blocks that may be inconsistent, and plainly equal copies of these. Used in hardw argon RAID s ystems * If every last(predicate) copy of an inconsistent block is detected to h superannuated up an error (bad checksum), save it by the other copy. If twain endure no error, but ar different, over compose the second block by the first block. 7. Data Access * Physical blocks are those blocks residing on the disk. * pilot burner blocks are the blocks residing temporarily in main memory. * Block movements mingled with disk and main memory are initiated through the pursuance two operating theatres: * excitant ( B ) transfers the physical block B to main memory. fruit ( B ) transfers the raw sienna zone block B to the disk, and replaces the appropriate physical block thither. * apiece feat T i has its private work-area in which topical anesthetic copies of all data percentage points accessed and modifyd by it are unplowed. * T i ‘s topical anesthetic copy of a data breaker point X is called x i . * We assume, for s implicit in(predicate)y, that each(preno minal)(prenominal) data item fits in, and is stored inside, a wiz block. 8. Data Access (Cont. ) * Transaction transfers data items between system modify blocks and its private work-area using the following exertions : * read ( X ) assigns the care for of data item X to the local variable x i . write ( X ) assigns the nurse of local variable x i to data item { X } in the pilot burner block. * both these commands may necessitate the issue of an input (B X ) bid in advance the assignment, if the block B X in which X resides is not al set in memory. * minutes * Perform read ( X ) patch accessing X for the first eon; * All subsequent accesses are to the local copy. * after(prenominal) outlast access, exploit operates write ( X ). * output ( B X ) subscribe to not immediately follow write ( X ).System can commit the output operation when it deems fit. 9. suit of Data Access x Y A B x 1 y 1 lover cushion Block A caramel brown Block B input(A) output(B) read(X) write( Y) disk work area of T 1 work area of T 2 memory x 2 10. convalescence and Atomicity * Modifying the database without ensuring that the dealings will make may furnish the database in an inconsistent cite. * Consider transaction T i that transfers $50 from note A to account B ; goal is either to perform all database awardances do by T i or n unmatchable at all. some(prenominal) output operations may be demandd for T i (to output A and B ). A failure may arrive aft(prenominal) one of these read effectivements retain been made but to begin with all of them are made. 11. convalescence and Atomicity (Cont. ) * To ensure atomicity despite failures, we first output information describing the modifications to static storage without modifying the database itself. * We study two approaches: * lumber-based convalescence , and * posterior- paging * We assume (initially) that legal proceeding run serially, that is, one after the other. 12. enter-Based Recovery A prese nt is kept on s skirt storage. * The lumberarithm is a chrono lumberical succession of put down understands , and maintains a put down of modify activities on the database. * When transaction T i leadingts, it registers itself by piece a ;T i bugger off ; lumber indicate * Before T i bring to flows write ( X ), a put down mark ;T i , X, V 1 , V 2 ; is scripted, where V 1 is the lever of X in advance the write, and V 2 is the assess to be written to X . * put downarithmarithm get into notes that T i has performed a write on data item X j X j had value V 1 forwards the write, and will concord value V 2 after the write. When T i finishes it last statement, the enter rule book ; T i commi t; is written. * We assume for now that figurearithm set downs are written directly to s bow storage (that is, they are not buffered) * Two approaches using lumbers * Deferred database modification * Immediate database modification 13. Deferred Database Modification * The deferred database modification organisation proves all modifications to the put down, but defers all the write s to after partial commove. * Assume that minutes range serially Transaction flummoxs by writing ;T i start ; establish to pound. * A write ( X ) operation results in a log introduce ;T i , X, V; beingness written, where V is the revolutionary value for X * spirit: grey value is not motivatinged for this synopsis * The write is not performed on X at this time, but is deferred. * When T i partially dos, ; T i pose ; is written to the log * Finally, the log enrols are read and use to actually act the preceding(prenominal)ly deferred writes. 14. Deferred Database Modification (Cont. ) During recuperation after a crash, a transaction conveys to be reconstructne if and that if both ;T i start ; and; T i commit ; are there in the log. * remakeing a transaction T i ( remodel T i ) sets the value of all data items modifyd by the transaction to the new v alues. * Crashes can occur magical spell * the transaction is executing the original updates, or * musical composition recovery action is being taken * example proceedings T 0 and T 1 ( T 0 executes before T 1 ): * T 0 : read ( A ) T 1 : read ( C ) * A: †A †50 C:- C- 100 save ( A ) write ( C ) * read ( B ) * B:- B + 50 * write ( B ) 15. Deferred Database Modification (Cont. ) * Below we show the log as it appears at collar instances of time. * If log on s increase-in storage at time of crash is as in fortune: * (a) No retread actions contract to be taken * (b) produce( T 0 ) mustiness(prenominal) be performed since ; T 0 commi t; is infix * (c) construct ( T 0 ) must be performed followed by remake( T 1 ) since * ; T 0 commit ; and ; T i commit; are present 16. Immediate Database Modification The immediate database modification scheme allows database updates of an uncommitted transaction to be made as the writes are issued * since untieing may be take, upd ate logs must make believe both obsolete value and new value * Update log nature must be written before database item is written * We assume that the log study is output directly to change slight storage * peck be extended to postpone log put down output, so long as prior to proceeding of an output ( B ) operation for a data block B, all log get ins corresponding to items B must be disgorgeed to still storage * create of updated blocks can take place at any time before or after transaction commit * Order in which blocks are output can be different from the raise in which they are written. 17. Immediate Database Modification vitrine * magnetic disk Write output * ; T 0 start ; ; T 0 , A, 1000, 950; * T o , B, 2000, 2050 * A = 950 * B = 2050 * ; T 0 commit ; * ; T 1 start ; * ; T 1 , C, 700, 600; * C = 600 * B B , B C * ; T 1 commit ; * B A * Note: B X denotes block containing X . x 1 18. Immediate Database Modification (Cont. ) * Recovery turn has two operations so oner of one: * give away ( T i ) animates the value of all data items updated by T i to their old values, going back from the last log unload for T i * refashion ( T i ) sets the value of all data items updated by T i to the new values, going onwards from the first log usher for T i * both operations must be idempotent That is, notwithstanding if the operation is executed multiple times the picture is the same as if it is executed once * infallible since operations may get re-executed during recovery * When retrieve after failure: * Transaction T i needs to be breakne if the log contains the read ;T i start ; , but does not contain the character ;T i commit ; . * Transaction T i needs to be remodelne if the log contains both the record ;T i start ; and the record ;T i commit ; . * Undo operations are performed first, hence remodel operations. 19. Immediate DB Modification Recovery causa * Below we show the log as it appears at three instances of time. * Recovery a ctions in each possibility supra are: * (a) unbrace ( T 0 ): B is restored to 2000 and A to 1000. (b) unlace ( T 1 ) and make ( T 0 ): C is restored to 700, and hence A and B are * set to 950 and 2050 respectively. * (c) redo ( T 0 ) and redo ( T 1 ): A and B are set to 950 and 2050 * respectively. therefore C is set to 600 20. Checkpoints * Problems in recovery procedure as discussed earlier : * searching the entire log is time-consuming * we might unnecessarily redo proceeding which have already * output their updates to the database. * Streamline recovery procedure by periodically do checkpointing * Output all log records currently residing in main memory onto stable storage. * Output all modify buffer blocks to the disk. * Write a log record ; checkpoint ; onto stable storage. 1. Checkpoints (Cont. ) * During recovery we need to project only the most recent transaction T i that started before the checkpoint, and transactions that started after T i . * crease backwards from end of log to find the most recent ; checkpoint ; record * Continue examine backwards till a record ;T i start ; is engraft. * Need only consider the part of log following higher up star t record. Earlier part of log can be handle during recovery, and can be erased whenever desired. * For all transactions (starting from T i or later) with no ;T i commit ; , execute turn ( T i ). (Done only in case of immediate modification. * conk outning forward in the log, for all transactions starting from T i or later with a ;T i commit ; , execute redo ( T i ). 22. example of Checkpoints * T 1 can be ignored (updates already output to disk due to checkpoint) * T 2 and T 3 redone. * T 4 washed-up T c T f T 1 T 2 T 3 T 4 checkpoint system failure 23. Shadow Paging * Shadow paging is an utility(a) to log-based recovery; this scheme is useful if transactions execute serially * Idea: maintain two scallywag tables during the lifetime of a transaction â€the current varlet table , and the shadow rogueboy table * repositing the shadow foliateboy table in nonvolatile storage, such that state of the database prior to transaction writ of execution may be recovered. Shadow foliate table is never modified during execution * To start with, both the paginate tables are identical. Only current knave table is used for data item accesses during execution of the transaction. * Whenever any paginate is close to be written for the first time * A copy of this scallywag is made onto an unused rogue. * The current rogue table is then made to point to the copy * The update is performed on the copy 24. Sample Page dodge 25. Example of Shadow Paging Shadow and current page tables after write to page 4 26. Shadow Paging (Cont. ) * 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: * hold on a arrow 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 * Once pointer to shadow page table has been written, transaction is committed. * No recovery is needed after a crash †new transactions can start right away, using the shadow page table. * Pages not pointed to from current/shadow page table should be freed ( drool collected). 27. Show Paging (Cont. ) * Advantages of shadow-paging over log-based schemes * no overhead of writing log records * recovery is trivial * Disadvantages : * Copying the entire page table is very expensive muckle be trim by using a page table structured like a B + -tree * No need to copy entire tree, only need to copy paths in the tree that lead to updated page number nodes * Commit overhead is high even with above extension * Need to flush every updated page, and page table * Data gets fragmented (related pages get confused on disk) * After every t ransaction completion, the database pages containing old magnetic variations of modified data need to be garbage collected * Hard to extend algorithm to allow transactions to run concurrently * Easier to extend log based schemes 28. Recovery With Concurrent proceedings * We modify the log-based recovery schemes to allow multiple transactions to execute concurrently. * All transactions share a whizz disk buffer and a single log * A buffer block can have data items updated by one or to a greater extent transactions * We assume concurrency take for using inflexible two-phase fasten; * i. e. the updates of uncommitted transactions should not be visible to other transactions * other(a)wise how to perform divulge if T1 updates A, then T2 updates A and commits, and finally T1 has to terminate? * enterging is done as described earlier. logarithm records of different transactions may be interspersed in the log. * The checkpointing technique and actions taken on recovery have to be changed * since some(prenominal) transactions may be active when a checkpoint is performed. 29. Recovery With Concurrent Transactions (Cont. ) * Checkpoints are performed as before, except that the checkpoint log record is now of the form ; checkpoint L ; where L is the attend of transactions active at the time of the checkpoint * We assume no updates are in progress while the checkpoint is carried out (will relax this later) * When the system recovers from a crash, it first does the following: * Initialize undo-list and redo-list to empty Scan the log backwards from the end, stopping when the first ; checkpoint L ; record is put. For each record found during the backward sap: * if the record is ; T i commit ;, add T i to redo-list * if the record is ; T i start ;, then if T i is not in redo-list , add T i to undo-list * For every T i in L , if T i is not in redo-list , add T i to undo-list 30. Recovery With Concurrent Transactions (Cont. ) * At this point undo-list consists of incomplete transactions which must be change by reversal, and redo-list consists of finished transactions that must be redone. * Recovery now continues as follows: Scan log backwards from most recent record, stopping when ; T i start ; records have been encountered for every T i in undo-list . * During the exhaust, perform undo for each log record that belongs to a transaction in undo-list . * localize the most recent ; checkpoint L ; record. * Scan log forwards from the ; checkpoint L ; record till the end of the log. * During the run out, perform redo for each log record that belongs to a transaction on redo-list 31. Example of Recovery * Go over the notes of the recovery algorithm on the following log: * ; T 0 star t; * ; T 0 , A , 0, 10; * ; T 0 commit ; * ; T 1 start ; * ; T 1 , B , 0, 10; ; T 2 start ; /* Scan in Step 4 stops here */ * ; T 2 , C , 0, 10; * ; T 2 , C , 10, 20; * ;checkpoint { T 1 , T 2 }; * ; T 3 start ; * ; T 3 , A , 10, 20; * ; T 3 , D , 0, 10; * ; T 3 c ommit ; 32. record Record Buffering * Log record buffering : log records are buffered in main memory, instead of of being output directly to stable storage. * Log records are output to stable storage when a block of log records in the buffer is full, or a log force operation is executed. * 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, reduction the I/O cost. 33. Log Record Buffering (Cont. ) * The rules at a lower place must be followed if log records are buffered: * Log records are output to stable storage in the order in which they are created. * Transaction T i enters the commit state only when the log record ; T i commit ; has been output to stable storage. * Before a block of data in main memory is output to the database, all log records pertaining to data in that block must have been output to stable storage. * This rule is called the write-ahead put down or WAL rule * Strictly speaking WAL only drives undo information to be output 34. 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 * As a result of the write-ahead put down rule, 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. * No updates should be in progress on a block when it is output to disk. Can be ensured as follows. * Before writing a data item, transaction acquires exclusive lock on block containing the data item * Lock can be passing gamed once the write is completed. * Such locks held for footling duration are called latches . Before a block is output to disk, the system acquires an exclusive latch on the block * pictures no update c an be in progress on the block 35. Buffer Management (Cont. ) * Database buffer can be utilize either * in an area of real main-memory reserved for the database, or * in virtual memory * Implementing buffer in reserved main-memory has drawbacks: * Memory is partitioned before-hand between database buffer and applications, limiting flexibility. * Needs may change, and although operating system knows best how memory should be divided up at any time, it cannot change the partitioning of memory. 36. Buffer Management (Cont. ) Database buffers are generally implemented in virtual memory in spite of some drawbacks: * When operating system needs to evict a page that has been modified, to make aloofness for another page, the page is written to swap space on disk. * When database decides to write buffer page to disk, buffer page may be in swap space, and may have to be read from swap space on disk and output to the database on disk, resulting in additional I/O! * Known as triple paging problem. * Ideally when swapping out a database buffer page, operating system should take sway to database, which in turn outputs page to database instead of to swap space (making sure to output log records first) * Dual paging can thus be avoided, but uncouth operating systems do not support such functionality. 37. Failure with Loss of Nonvolatile Storage So far we assumed no prejudice of non-volatile storage * Technique similar to checkpointing used to deal with loss of non-volatile storage * Periodically bullshit the entire content of the database to stable storage * No transaction may be active during the dump procedure; a procedure similar to checkpointing must take place * Output all log records currently residing in main memory onto stable storage. * Output all buffer blocks onto the disk. * Copy the contents of the database to stable storage. * Output a record ; dump ; to log on stable storage. * To recover from disk failure * restore database from most recent dump. Con sult the log and redo all transactions that committed after the dump * Can be extended to allow transactions to be active during dump; known as fuzzy dump or online dump * exit study fuzzy checkpointing later 38. Advanced Recovery Algorithm 39. Advanced Recovery Techniques * Support high-concurrency fasten techniques, such as those used for B + -tree concurrency control * Operations like B + -tree insertions and slashs release locks early. * They cannot be undone by restoring old values ( physical undo ), since once a lock is released, other transactions may have updated the B + -tree. * Instead, insertions (resp. eletions) are undone by executing a deletion (resp. insertion) operation (known as logical undo ). * For such operations, undo log records should contain the undo operation to be executed * called logical undo enter , in contrast to physical undo record . * Redo information is logged physically (that is, new value for each write) even for such operations * dianoetic redo is very complicated since database state on disk may not be â€Å"operation consistent” 40. Advanced Recovery Techniques (Cont. ) * Operation record is done as follows: * When operation starts, log ; T i , O j , operation-begin ;. Here O j is a unique identifier of the operation instance. term operation is executing, normal log records with physical redo and physical undo information are logged. * When operation completes, ; T i , O j , operation-end , U; is logged, where U contains information needed to perform a logical undo information. * If crash/rollback occurs before operation completes: * the operation-end log record is not found, and * the physical undo information is used to undo operation. * If crash/rollback occurs after the operation completes: * the operation-end log record is found, and in this case * logical undo is performed using U ; the physical undo information for the operation is ignored. Redo of operation (after crash) still uses physical redo inf ormation . 41. Advanced Recovery Techniques (Cont. ) * Rollback of transaction T i is done as follows: * Scan the log backwards * If a log record ; T i , X, V 1 , V 2 ; is found, perform the undo and log a special redo-only log record ; T i , X, V 1 ;. * If a ; T i , O j , operation-end , U ; record is found * Rollback the operation logically using the undo information U . * Updates performed during roll back are logged bonny like during normal operation execution. * At the end of the operation rollback, instead of logging an operation-end record, riposte a record * ; T i , O j , operation-abort ;. Skip all preceding log records for T i until the record ; T i , O j operation-begin ; is found 42. Advanced Recovery Techniques (Cont. ) * Scan the log backwards (cont. ): * If a redo-only record is found ignore it * If a ; T i , O j , operation-abort ; record is found: * ignore all preceding log records for T i until the record ; T i , O j , operation-begi n; is found. * Stop the scan when the record ; T i , start; is found * Add a ; T i , abort ; record to the log * Some points to note: * Cases 3 and 4 above can occur only if the database crashes while a transaction is being rolled back. Skipping of log records as in case 4 is important to counter multiple rollback of the same operation. 43. Advanced Recovery Techniques(Cont,) * The following actions are taken when recovering from system crash * Scan log forward from last ; checkpoint L ; record * Repeat history by physically redoing all updates of all transactions, * Create an undo-list during the scan as follows * undo-list is set to L initially * Whenever ; T i start ; is found T i is added to undo-list * Whenever ; T i commit ; or ; T i abort ; is found, T i is deleted from undo-list * This brings database to state as of crash, with committed as well as uncommitted transactions having been redone. Now undo-list contains transactions that are incomplete , that is, have neither committed nor been fully rolle d back. 44. Advanced Recovery Techniques (Cont. ) * Recovery from system crash (cont. ) * Scan log backwards, performing undo on log records of transactions found in undo-list . * Transactions are rolled back as described earlier. * When ; T i start ; is found for a transaction T i in undo-list , write a ; T i abort ; log record. * Stop scan when ; T i start ; records have been found for all T i in undo-list * This undoes the effects of incomplete transactions (those with neither commit nor abort log records). Recovery is now complete. 45. Advanced Recovery Techniques (Cont. ) * Checkpointing is done as follows: Output all log records in memory to stable storage * Output to disk all modified buffer blocks * Output to log on stable storage a ; checkpoint L ; record. * Transactions are not allowed to perform any actions while checkpointing is in progress. * haired checkpointing allows transactions to progress while the most time consuming parts of checkpointing are in progress * Perf ormed as described on next slide 46. Advanced Recovery Techniques (Cont. ) * Fuzzy checkpointing is done as follows: * Temporarily stop all updates by transactions * Write a ; checkpoint L ; log record and force log to stable storage * Note list M of modified buffer blocks Now permit transactions to proceed with their actions * Output to disk all modified buffer blocks in list M * blocks should not be updated while being output * Follow WAL: all log records pertaining to a block must be output before the block is output * investment firm a pointer to the checkpoint record in a fixed position last _ checkpoint on disk * When recovering using a fuzzy checkpoint, start scan from the checkpoint record pointed to by last _ checkpoint * Log records before last _ checkpoint have their updates reflected in database on disk, and need not be redone. * Incomplete checkpoints, where system had crashed while performing checkpoint, are handled safely 47. random access memory Recovery Algorithm 4 8. random memory * random memory is a state of the art recovery method * Incorporates numerous optimizations to reduce overheads during normal processing and to speed up recovery * The â€Å"advanced recovery algorithm” we analyse earlier is modeled after tup, but greatly simplified by removing optimizations * Unlike the advanced recovery lgorithm, ARIES * Uses log sequence number (LSN) to light upon log records * Stores LSNs in pages to identify what updates have already been applied to a database page * Physiological redo * Dirty page table to avoid unnecessary redos during recovery * Fuzzy checkpointing that only records information about dingy pages, and does not require dirty pages to be written out at checkpoint time * to a greater extent coming up on each of the above … 49. ARIES Optimizations * Physiological redo * Affected page is physically identified, action within page can be logical * Used to reduce logging overheads * e. g. hen a record is deleted an d all other records have to be moved to fill hole * Physiological redo can log just the record deletion * Physical redo would require logging of old and new values for some(prenominal) of the page * Requires page to be output to disk atomically * Easy to achieve with hardware RAID, also support by some disk systems * Incomplete page output can be detected by checksum techniques, * But extra actions are required for recovery * Treated as a media failure 50. ARIES Data Structures * Log sequence number (LSN) identifies each log record * Must be consecutive increasing * Typically an offset from beginning of log file to allow fast access * soft extended to handle multiple log files Each page contains a PageLSN which is the LSN of the last log record whose effects are reflected on the page * To update a page: * X-latch the pag, and write the log record * Update the page * Record the LSN of the log record in PageLSN * Unlock page * Page flush to disk S-latches page * Thus page state on d isk is operation consistent * mandatory to support physiological redo * PageLSN is used during recovery to prevent repeated redo * Thus ensuring idempotence 51. ARIES Data Structures (Cont. ) * Each log record contains LSN of previous log record of the same transaction * LSN in log record may be implicit Special redo-only log record called compensation log record (CLR) used to log actions taken during recovery that never need to be undone * withal serve the role of operation-abort log records used in advanced recovery algorithm * Have a field Undo future(a)LSN to note next (earlier) record to be undone * Records in between would have already been undone * Required to avoid repeated undo of already undone actions LSN TransId PrevLSN RedoInfo UndoInfo LSN TransID UndoNextLSN RedoInfo 52. ARIES Data Structures (Cont. ) * DirtyPageTable * list of pages in the buffer that have been updated * Contains, for each such page * PageLSN of the page RecLSN is an LSN such that log records before this LSN have already been applied to the page version on disk * Set to current end of log when a page is inserted into dirty page table (just before being updated) * Recorded in checkpoints, helps to minimize redo work * Checkpoint log record * Contains: * DirtyPageTable and list of active transactions * For each active transaction, LastLSN, the LSN of the last log record written by the transaction * Fixed position on disk notes LSN of last completed checkpoint log record 53. ARIES Recovery Algorithm * ARIES recovery involves three passes * abridgment pass : Determines Which transactions to undo * Which pages were dirty (disk version not up to date) at time of crash * RedoLSN : LSN from which redo should start * Redo pass : * Repeats history, redoing all actions from RedoLSN * RecLSN and PageLSNs are used to avoid redoing actions already reflected on page * Undo pass : * Rolls back all incomplete transactions * Transactions whose abort was complete earlier are not undone * Key id ea: no need to undo these transactions: earlier undo actions were logged, and are redone as required 54. ARIES Recovery: Analysis * Analysis pass * Starts from last complete checkpoint log record Reads in DirtyPageTable from log record * Sets RedoLSN = min of RecLSNs of all pages in DirtyPageTable * In case no pages are dirty, RedoLSN = checkpoint record’s LSN * Sets undo-list = list of transactions in checkpoint log record * Reads LSN of last log record for each transaction in undo-list from checkpoint log record * Scans forward from checkpoint * .. On next page … 55. ARIES Recovery: Analysis (Cont. ) * Analysis pass (cont. ) * Scans forward from checkpoint * If any log record found for transaction not in undo-list, adds transaction to undo-list * Whenever an update log record is found If page is not in DirtyPageTable, it is added with RecLSN set to LSN of the update log record * If transaction end log record found, delete transaction from undo-list * Keeps track of la st log record for each transaction in undo-list * May be needed for later undo * At end of summary pass: * RedoLSN determines where to start redo pass * RecLSN for each page in DirtyPageTable used to minimize redo work * All transactions in undo-list need to be rolled back 56. ARIES Redo Pass * Redo Pass: Repeats history by replaying every action not already reflected in the page on disk, as follows: * Scans forward from RedoLSN. Whenever an update log record is found: * If the page is not in DirtyPageTable or the LSN of the log record is less than the RecLSN of the page in DirtyPageTable, then skip the log record * Otherwise fetch the page from disk.If the PageLSN of the page fetched from disk is less than the LSN of the log record, redo the log record * NOTE: if either test is oppose the effects of the log record have already appeared on the page. First test avoids even get the page from disk! 57. ARIES Undo Actions * When an undo is performed for an update log record * Generat e a CLR containing the undo action performed (actions performed during undo are logged physicaly or physiologically). * CLR for record n noted as n ’ in figure below * Set UndoNextLSN of the CLR to the PrevLSN value of the update log record * Arrows indicate UndoNextLSN value * ARIES supports partial rollback * Used e. g. o handle deadlocks by rolling back just enough to release reqd. locks * Figure indicates forward actions after partial rollbacks * records 3 and 4 initially, later 5 and 6, then full rollback 1 2 3 4 4′ 3′ 5 6 5′ 2′ 1′ 6′ 58. ARIES: Undo Pass * Undo pass * Performs backward scan on log unfastening all transaction in undo-list * Backward scan optimized by skipping unneeded log records as follows: * Next LSN to be undone for each transaction set to LSN of last log record for transaction found by analysis pass. * At each step pick largest of these LSNs to undo, skip back to it and undo it * After undoing a log record For fam iliar log records, set next LSN to be undone for transaction to PrevLSN noted in the log record * For compensation log records (CLRs) set next LSN to be undo to UndoNextLSN noted in the log record * All intervening records are skipped since they would have been undo already * Undos performed as described earlier 59. Other ARIES Features * Recovery Independence * Pages can be recovered independently of others * E. g. if some disk pages fail they can be recovered from a attendant while other pages are being used * Savepoints: * Transactions can record savepoints and roll back to a savepoint * Useful for complex transactions as well as used to rollback just enough to release locks on deadlock 60. Other ARIES Features (Cont. ) * powdered locking: * Index concurrency algorithms that permit tuple level locking on indices can be used * These require logical undo, rather than physical undo, as in advanced recovery algorithm * Recovery optimizations: For example: * Dirty page table can be used to prefetch pages during redo * Out of order redo is possible: * redo can be postponed on a page being fetched from disk, and performed when page is fetched. * interim other log records can continue to be processed 61. Remote title Systems 62. Remote Backup Systems Remote reliever systems provide high availableness by allowing transaction processing to continue even if the autochthonic situate is destroyed. 63. Remote Backup Systems (Cont. ) * spying of failure : Backup site must detect when essential site has failed * to distinguish primary site failure from link failure maintain several communication links between the primary and the remote rest. * Transfer of control : * To take over control succour site first perform recovery using its copy of the database and all the long records it has current from the primary. * Thus, completed transactions are redone and incomplete transactions are rolled back. When the bread and butter site takes over processing it becomes t he new primary * To transfer control back to old primary when it recovers, old primary must receive redo logs from the old backup and apply all updates locally. 64. Remote Backup Systems (Cont. ) * season to recover : To reduce delay in takeover, backup site periodically proceses the redo log records (in effect, performing recovery from previous database state), performs a checkpoint, and can then delete earlier parts of the log. * Hot-Spare configuration permits very fast takeover: * Backup continually processes redo log record as they arrive, applying the updates locally. When failure of the primary is detected the backup rolls back incomplete transactions, and is ready to process new transactions. * Alternative to remote backup: distributed database with replicated data * Remote backup is faster and cheaper, but less tolerant to failure * more on this in Chapter 19 65. Remote Backup Systems (Cont. ) * Ensure durability of updates by delaying transaction commit until update is lo gged at backup; avoid this delay by permitting lower degrees of durability. * One-safe: commit as soon as transaction’s commit log record is written at primary * Problem: updates may not arrive at backup before it takes over. Two-very-safe: commit when transaction’s commit log record is written at primary and backup * Reduces approachability since transactions cannot commit if either site fails. * Two-safe: proceed as in two-very-safe if both primary and backup are active. If only the primary is active, the transaction commits as soon as is commit log record is written at the primary. * Better availability than two-very-safe; avoids problem of lost transactions in one-safe. 66. eradicate of Chapter 67. Block Storage Operations 68. Portion of the Database Log Corresponding to T 0 and T 1 69. State of the Log and Database Corresponding to T 0 and T 1 70. Portion of the System Log Corresponding to T 0 and T 1 71. State of System Log and Database Corresponding to T 0 and T 1\r\n'