Chapter 22: Distributed Databases

Chapter 22: Distributed Databases

اسلاید 1: Chapter 22: Distributed Databases

اسلاید 2: Chapter 22: Distributed DatabasesHeterogeneous and Homogeneous DatabasesDistributed Data StorageDistributed TransactionsCommit ProtocolsConcurrency Control in Distributed DatabasesAvailabilityDistributed Query ProcessingHeterogeneous Distributed DatabasesDirectory Systems

اسلاید 3: Distributed Database SystemA distributed database system consists of loosely coupled sites that share no physical componentDatabase systems that run on each site are independent of each otherTransactions may access data at one or more sites

اسلاید 4: Homogeneous Distributed DatabasesIn a homogeneous distributed databaseAll sites have identical software Are aware of each other and agree to cooperate in processing user requests.Each site surrenders part of its autonomy in terms of right to change schemas or softwareAppears to user as a single systemIn a heterogeneous distributed databaseDifferent sites may use different schemas and softwareDifference in schema is a major problem for query processingDifference in software is a major problem for transaction processingSites may not be aware of each other and may provide only limited facilities for cooperation in transaction processing

اسلاید 5: Distributed Data StorageAssume relational data modelReplicationSystem maintains multiple copies of data, stored in different sites, for faster retrieval and fault tolerance.FragmentationRelation is partitioned into several fragments stored in distinct sitesReplication and fragmentation can be combinedRelation is partitioned into several fragments: system maintains several identical replicas of each such fragment.

اسلاید 6: Data ReplicationA relation or fragment of a relation is replicated if it is stored redundantly in two or more sites.Full replication of a relation is the case where the relation is stored at all sites.Fully redundant databases are those in which every site contains a copy of the entire database.

اسلاید 7: Data Replication (Cont.)Advantages of ReplicationAvailability: failure of site containing relation r does not result in unavailability of r is replicas exist.Parallelism: queries on r may be processed by several nodes in parallel.Reduced data transfer: relation r is available locally at each site containing a replica of r.Disadvantages of ReplicationIncreased cost of updates: each replica of relation r must be updated.Increased complexity of concurrency control: concurrent updates to distinct replicas may lead to inconsistent data unless special concurrency control mechanisms are implemented.One solution: choose one copy as primary copy and apply concurrency control operations on primary copy

اسلاید 8: Data FragmentationDivision of relation r into fragments r1, r2, …, rn which contain sufficient information to reconstruct relation r.Horizontal fragmentation: each tuple of r is assigned to one or more fragmentsVertical fragmentation: the schema for relation r is split into several smaller schemasAll schemas must contain a common candidate key (or superkey) to ensure lossless join property.A special attribute, the tuple-id attribute may be added to each schema to serve as a candidate key.Example : relation account with following schemaAccount = (branch_name, account_number, balance )

اسلاید 9: Horizontal Fragmentation of account Relationbranch_nameaccount_numberbalanceHillsideHillsideHillsideA-305A-226A-15550033662account1 = branch_name=“Hillside” (account )branch_nameaccount_numberbalanceValleyviewValleyviewValleyviewValleyviewA-177A-402A-408A-639205100001123750account2 = branch_name=“Valleyview” (account )

اسلاید 10: Vertical Fragmentation of employee_info Relationbranch_namecustomer_nametuple_idHillsideHillsideValleyviewValleyviewHillsideValleyviewValleyviewLowmanCampCampKahnKahnKahnGreendeposit1 = branch_name, customer_name, tuple_id (employee_info )1234567account_numberbalancetuple_id500336205100006211237501234567A-305A-226A-177A-402A-155A-408A-639deposit2 = account_number, balance, tuple_id (employee_info )

اسلاید 11: Advantages of FragmentationHorizontal:allows parallel processing on fragments of a relationallows a relation to be split so that tuples are located where they are most frequently accessedVertical: allows tuples to be split so that each part of the tuple is stored where it is most frequently accessedtuple-id attribute allows efficient joining of vertical fragmentsallows parallel processing on a relationVertical and horizontal fragmentation can be mixed.Fragments may be successively fragmented to an arbitrary depth.

اسلاید 12: Data TransparencyData transparency: Degree to which system user may remain unaware of the details of how and where the data items are stored in a distributed systemConsider transparency issues in relation to:Fragmentation transparencyReplication transparencyLocation transparency

اسلاید 13: Naming of Data Items - Criteria1. Every data item must have a system-wide unique name.2. It should be possible to find the location of data items efficiently.3. It should be possible to change the location of data items transparently.4. Each site should be able to create new data items autonomously.

اسلاید 14: Centralized Scheme - Name ServerStructure:name server assigns all nameseach site maintains a record of local data itemssites ask name server to locate non-local data itemsAdvantages:satisfies naming criteria 1-3Disadvantages:does not satisfy naming criterion 4name server is a potential performance bottleneckname server is a single point of failure

اسلاید 15: Use of AliasesAlternative to centralized scheme: each site prefixes its own site identifier to any name that it generates i.e., site 17.account.Fulfills having a unique identifier, and avoids problems associated with central control.However, fails to achieve network transparency.Solution: Create a set of aliases for data items; Store the mapping of aliases to the real names at each site.The user can be unaware of the physical location of a data item, and is unaffected if the data item is moved from one site to another.

اسلاید 16: Distributed TransactionsTransaction may access data at several sites.Each site has a local transaction manager responsible for:Maintaining a log for recovery purposesParticipating in coordinating the concurrent execution of the transactions executing at that site.Each site has a transaction coordinator, which is responsible for:Starting the execution of transactions that originate at the site.Distributing subtransactions at appropriate sites for execution.Coordinating the termination of each transaction that originates at the site, which may result in the transaction being committed at all sites or aborted at all sites.

اسلاید 17: Transaction System Architecture

اسلاید 18: System Failure ModesFailures unique to distributed systems:Failure of a site.Loss of massagesHandled by network transmission control protocols such as TCP-IPFailure of a communication linkHandled by network protocols, by routing messages via alternative linksNetwork partitionA network is said to be partitioned when it has been split into two or more subsystems that lack any connection between themNote: a subsystem may consist of a single node Network partitioning and site failures are generally indistinguishable.

اسلاید 19: Commit ProtocolsCommit protocols are used to ensure atomicity across sitesa transaction which executes at multiple sites must either be committed at all the sites, or aborted at all the sites.not acceptable to have a transaction committed at one site and aborted at anotherThe two-phase commit (2PC) protocol is widely used The three-phase commit (3PC) protocol is more complicated and more expensive, but avoids some drawbacks of two-phase commit protocol. This protocol is not used in practice.

اسلاید 20: Two Phase Commit Protocol (2PC)Assumes fail-stop model – failed sites simply stop working, and do not cause any other harm, such as sending incorrect messages to other sites.Execution of the protocol is initiated by the coordinator after the last step of the transaction has been reached.The protocol involves all the local sites at which the transaction executedLet T be a transaction initiated at site Si, and let the transaction coordinator at Si be Ci

اسلاید 21: Phase 1: Obtaining a DecisionCoordinator asks all participants to prepare to commit transaction Ti.Ci adds the records <prepare T> to the log and forces log to stable storagesends prepare T messages to all sites at which T executedUpon receiving message, transaction manager at site determines if it can commit the transactionif not, add a record <no T> to the log and send abort T message to Ciif the transaction can be committed, then:add the record <ready T> to the logforce all records for T to stable storagesend ready T message to Ci

اسلاید 22: Phase 2: Recording the DecisionT can be committed of Ci received a ready T message from all the participating sites: otherwise T must be aborted.Coordinator adds a decision record, <commit T> or <abort T>, to the log and forces record onto stable storage. Once the record stable storage it is irrevocable (even if failures occur)Coordinator sends a message to each participant informing it of the decision (commit or abort)Participants take appropriate action locally.

اسلاید 23: Handling of Failures - Site FailureWhen site Si recovers, it examines its log to determine the fate oftransactions active at the time of the failure.Log contain <commit T> record: site executes redo (T)Log contains <abort T> record: site executes undo (T)Log contains <ready T> record: site must consult Ci to determine the fate of T.If T committed, redo (T)If T aborted, undo (T)The log contains no control records concerning T replies that Sk failed before responding to the prepare T message from Ci since the failure of Sk precludes the sending of such a response C1 must abort TSk must execute undo (T)

اسلاید 24: Handling of Failures- Coordinator FailureIf coordinator fails while the commit protocol for T is executing then participating sites must decide on T’s fate:If an active site contains a <commit T> record in its log, then T must be committed.If an active site contains an <abort T> record in its log, then T must be aborted.If some active participating site does not contain a <ready T> record in its log, then the failed coordinator Ci cannot have decided to commit T. Can therefore abort T.If none of the above cases holds, then all active sites must have a <ready T> record in their logs, but no additional control records (such as <abort T> of <commit T>). In this case active sites must wait for Ci to recover, to find decision.Blocking problem : active sites may have to wait for failed coordinator to recover.

اسلاید 25: Handling of Failures - Network PartitionIf the coordinator and all its participants remain in one partition, the failure has no effect on the commit protocol.If the coordinator and its participants belong to several partitions:Sites that are not in the partition containing the coordinator think the coordinator has failed, and execute the protocol to deal with failure of the coordinator.No harm results, but sites may still have to wait for decision from coordinator.The coordinator and the sites are in the same partition as the coordinator think that the sites in the other partition have failed, and follow the usual commit protocol.Again, no harm results

اسلاید 26: Recovery and Concurrency ControlIn-doubt transactions have a <ready T>, but neither a <commit T>, nor an <abort T> log record.The recovering site must determine the commit-abort status of such transactions by contacting other sites; this can slow and potentially block recovery.Recovery algorithms can note lock information in the log.Instead of <ready T>, write out <ready T, L> L = list of locks held by T when the log is written (read locks can be omitted).For every in-doubt transaction T, all the locks noted in the <ready T, L> log record are reacquired.After lock reacquisition, transaction processing can resume; the commit or rollback of in-doubt transactions is performed concurrently with the execution of new transactions.

اسلاید 27: Alternative Models of Transaction ProcessingNotion of a single transaction spanning multiple sites is inappropriate for many applicationsE.g. transaction crossing an organizational boundaryNo organization would like to permit an externally initiated transaction to block local transactions for an indeterminate periodAlternative models carry out transactions by sending messagesCode to handle messages must be carefully designed to ensure atomicity and durability properties for updatesIsolation cannot be guaranteed, in that intermediate stages are visible, but code must ensure no inconsistent states result due to concurrency Persistent messaging systems are systems that provide transactional properties to messages Messages are guaranteed to be delivered exactly onceWill discuss implementation techniques later

اسلاید 28: Alternative Models (Cont.)Motivating example: funds transfer between two banksTwo phase commit would have the potential to block updates on the accounts involved in funds transferAlternative solution:Debit money from source account and send a message to other siteSite receives message and credits destination accountMessaging has long been used for distributed transactions (even before computers were invented!)Atomicity issue once transaction sending a message is committed, message must guaranteed to be deliveredGuarantee as long as destination site is up and reachable, code to handle undeliverable messages must also be available e.g. credit money back to source account. If sending transaction aborts, message must not be sent

اسلاید 29: Error Conditions with Persistent MessagingCode to handle messages has to take care of variety of failure situations (even assuming guaranteed message delivery)E.g. if destination account does not exist, failure message must be sent back to source siteWhen failure message is received from destination site, or destination site itself does not exist, money must be deposited back in source accountProblem if source account has been closed get humans to take care of problemUser code executing transaction processing using 2PC does not have to deal with such failuresThere are many situations where extra effort of error handling is worth the benefit of absence of blockingE.g. pretty much all transactions across organizations

اسلاید 30: Persistent Messaging and WorkflowsWorkflows provide a general model of transactional processing involving multiple sites and possibly human processing of certain stepsE.g. when a bank receives a loan application, it may need toContact external credit-checking agenciesGet approvals of one or more managers and then respond to the loan applicationWe study workflows in Chapter 25Persistent messaging forms the underlying infrastructure for workflows in a distributed environment

اسلاید 31: Concurrency ControlModify concurrency control schemes for use in distributed environment.We assume that each site participates in the execution of a commit protocol to ensure global transaction automicity.We assume all replicas of any item are updated Will see how to relax this in case of site failures later

اسلاید 32: Single-Lock-Manager ApproachSystem maintains a single lock manager that resides in a single chosen site, say Si When a transaction needs to lock a data item, it sends a lock request to Si and lock manager determines whether the lock can be granted immediatelyIf yes, lock manager sends a message to the site which initiated the requestIf no, request is delayed until it can be granted, at which time a message is sent to the initiating site

اسلاید 33: Single-Lock-Manager Approach (Cont.)The transaction can read the data item from any one of the sites at which a replica of the data item resides.Writes must be performed on all replicas of a data itemAdvantages of scheme:Simple implementationSimple deadlock handlingDisadvantages of scheme are:Bottleneck: lock manager site becomes a bottleneckVulnerability: system is vulnerable to lock manager site failure.

اسلاید 34: Distributed Lock ManagerIn this approach, functionality of locking is implemented by lock managers at each siteLock managers control access to local data itemsBut special protocols may be used for replicasAdvantage: work is distributed and can be made robust to failuresDisadvantage: deadlock detection is more complicatedLock managers cooperate for deadlock detectionMore on this laterSeveral variants of this approachPrimary copyMajority protocolBiased protocolQuorum consensus

اسلاید 35: Primary CopyChoose one replica of data item to be the primary copy. Site containing the replica is called the primary site for that data itemDifferent data items can have different primary sitesWhen a transaction needs to lock a data item Q, it requests a lock at the primary site of Q.Implicitly gets lock on all replicas of the data itemBenefitConcurrency control for replicated data handled similarly to unreplicated data - simple implementation.DrawbackIf the primary site of Q fails, Q is inaccessible even though other sites containing a replica may be accessible.

اسلاید 36: Majority ProtocolLocal lock manager at each site administers lock and unlock requests for data items stored at that site.When a transaction wishes to lock an unreplicated data item Q residing at site Si, a message is sent to Si ‘s lock manager.If Q is locked in an incompatible mode, then the request is delayed until it can be granted.When the lock request can be granted, the lock manager sends a message back to the initiator indicating that the lock request has been granted.

اسلاید 37: Majority Protocol (Cont.)In case of replicated dataIf Q is replicated at n sites, then a lock request message must be sent to more than half of the n sites in which Q is stored.The transaction does not operate on Q until it has obtained a lock on a majority of the replicas of Q.When writing the data item, transaction performs writes on all replicas.BenefitCan be used even when some sites are unavailabledetails on how handle writes in the presence of site failure laterDrawbackRequires 2(n/2 + 1) messages for handling lock requests, and (n/2 + 1) messages for handling unlock requests.Potential for deadlock even with single item - e.g., each of 3 transactions may have locks on 1/3rd of the replicas of a data.

اسلاید 38: Biased ProtocolLocal lock manager at each site as in majority protocol, however, requests for shared locks are handled differently than requests for exclusive locks.Shared locks. When a transaction needs to lock data item Q, it simply requests a lock on Q from the lock manager at one site containing a replica of Q.Exclusive locks. When transaction needs to lock data item Q, it requests a lock on Q from the lock manager at all sites containing a replica of Q.Advantage - imposes less overhead on read operations.Disadvantage - additional overhead on writes

اسلاید 39: Quorum Consensus ProtocolA generalization of both majority and biased protocolsEach site is assigned a weight.Let S be the total of all site weightsChoose two values read quorum Qr and write quorum QwSuch that Qr + Qw > S and 2 * Qw > SQuorums can be chosen (and S computed) separately for each item Each read must lock enough replicas that the sum of the site weights is >= QrEach write must lock enough replicas that the sum of the site weights is >= QwFor now we assume all replicas are writtenExtensions to allow some sites to be unavailable described later

اسلاید 40: TimestampingTimestamp based concurrency-control protocols can be used in distributed systemsEach transaction must be given a unique timestampMain problem: how to generate a timestamp in a distributed fashionEach site generates a unique local timestamp using either a logical counter or the local clock.Global unique timestamp is obtained by concatenating the unique local timestamp with the unique identifier.

اسلاید 41: Timestamping (Cont.)A site with a slow clock will assign smaller timestampsStill logically correct: serializability not affectedBut: “disadvantages” transactionsTo fix this problemDefine within each site Si a logical clock (LCi), which generates the unique local timestampRequire that Si advance its logical clock whenever a request is received from a transaction Ti with timestamp < x,y> and x is greater that the current value of LCi.In this case, site Si advances its logical clock to the value x + 1.

اسلاید 42: Replication with Weak ConsistencyMany commercial databases support replication of data with weak degrees of consistency (I.e., without a guarantee of serializabiliy)E.g.: master-slave replication: updates are performed at a single “master” site, and propagated to “slave” sites. Propagation is not part of the update transaction: its is decoupledMay be immediately after transaction commitsMay be periodicData may only be read at slave sites, not updatedNo need to obtain locks at any remote siteParticularly useful for distributing informationE.g. from central office to branch-office Also useful for running read-only queries offline from the main database

اسلاید 43: Replication with Weak Consistency (Cont.)Replicas should see a transaction-consistent snapshot of the databaseThat is, a state of the database reflecting all effects of all transactions up to some point in the serialization order, and no effects of any later transactions. E.g. Oracle provides a create snapshot statement to create a snapshot of a relation or a set of relations at a remote sitesnapshot refresh either by recomputation or by incremental updateAutomatic refresh (continuous or periodic) or manual refresh

اسلاید 44: Multimaster and Lazy ReplicationWith multimaster replication (also called update-anywhere replication) updates are permitted at any replica, and are automatically propagated to all replicasBasic model in distributed databases, where transactions are unaware of the details of replication, and database system propagates updates as part of the same transactionCoupled with 2 phase commitMany systems support lazy propagation where updates are transmitted after transaction commitsAllows updates to occur even if some sites are disconnected from the network, but at the cost of consistency

اسلاید 45: Deadlock HandlingConsider the following two transactions and history, with item X and transaction T1 at site 1, and item Y and transaction T2 at site 2:T1: write (X)write (Y)T2: write (Y)write (X)X-lock on Xwrite (X)X-lock on Ywrite (Y)wait for X-lock on XWait for X-lock on YResult: deadlock which cannot be detected locally at either site

اسلاید 46: Centralized ApproachA global wait-for graph is constructed and maintained in a single site; the deadlock-detection coordinatorReal graph: Real, but unknown, state of the system.Constructed graph:Approximation generated by the controller during the execution of its algorithm .the global wait-for graph can be constructed when:a new edge is inserted in or removed from one of the local wait-for graphs.a number of changes have occurred in a local wait-for graph.the coordinator needs to invoke cycle-detection.If the coordinator finds a cycle, it selects a victim and notifies all sites. The sites roll back the victim transaction.

اسلاید 47: Local and Global Wait-For GraphsLocalGlobal

اسلاید 48: Example Wait-For Graph for False CyclesInitial state:

اسلاید 49: False Cycles (Cont.)Suppose that starting from the state shown in figure,1. T2 releases resources at S1 resulting in a message remove T1  T2 message from the Transaction Manager at site S1 to the coordinator)2. And then T2 requests a resource held by T3 at site S2 resulting in a message insert T2  T3 from S2 to the coordinatorSuppose further that the insert message reaches before the delete message this can happen due to network delaysThe coordinator would then find a false cycle T1  T2  T3  T1The false cycle above never existed in reality.False cycles cannot occur if two-phase locking is used.

اسلاید 50: Unnecessary RollbacksUnnecessary rollbacks may result when deadlock has indeed occurred and a victim has been picked, and meanwhile one of the transactions was aborted for reasons unrelated to the deadlock.Unnecessary rollbacks can result from false cycles in the global wait-for graph; however, likelihood of false cycles is low.

اسلاید 51: AvailabilityHigh availability: time for which system is not fully usable should be extremely low (e.g. 99.99% availability) Robustness: ability of system to function spite of failures of componentsFailures are more likely in large distributed systemsTo be robust, a distributed system must Detect failuresReconfigure the system so computation may continueRecovery/reintegration when a site or link is repairedFailure detection: distinguishing link failure from site failure is hard(partial) solution: have multiple links, multiple link failure is likely a site failure

اسلاید 52: ReconfigurationReconfiguration:Abort all transactions that were active at a failed siteMaking them wait could interfere with other transactions since they may hold locks on other sitesHowever, in case only some replicas of a data item failed, it may be possible to continue transactions that had accessed data at a failed site (more on this later) If replicated data items were at failed site, update system catalog to remove them from the list of replicas. This should be reversed when failed site recovers, but additional care needs to be taken to bring values up to dateIf a failed site was a central server for some subsystem, an election must be held to determine the new serverE.g. name server, concurrency coordinator, global deadlock detector

اسلاید 53: Reconfiguration (Cont.)Since network partition may not be distinguishable from site failure, the following situations must be avoidedTwo ore more central servers elected in distinct partitionsMore than one partition updates a replicated data itemUpdates must be able to continue even if some sites are downSolution: majority based approachAlternative of “read one write all available” is tantalizing but causes problems

اسلاید 54: Majority-Based ApproachThe majority protocol for distributed concurrency control can be modified to work even if some sites are unavailableEach replica of each item has a version number which is updated when the replica is updated, as outlined belowA lock request is sent to at least ½ the sites at which item replicas are stored and operation continues only when a lock is obtained on a majority of the sitesRead operations look at all replicas locked, and read the value from the replica with largest version numberMay write this value and version number back to replicas with lower version numbers (no need to obtain locks on all replicas for this task)

اسلاید 55: Majority-Based ApproachMajority protocol (Cont.)Write operations find highest version number like reads, and set new version number to old highest version + 1Writes are then performed on all locked replicas and version number on these replicas is set to new version numberFailures (network and site) cause no problems as long as Sites at commit contain a majority of replicas of any updated data itemsDuring reads a majority of replicas are available to find version numbersSubject to above, 2 phase commit can be used to update replicasNote: reads are guaranteed to see latest version of data itemReintegration is trivial: nothing needs to be doneQuorum consensus algorithm can be similarly extended

اسلاید 56: Read One Write All (Available)Biased protocol is a special case of quorum consensusAllows reads to read any one replica but updates require all replicas to be available at commit time (called read one write all)Read one write all available (ignoring failed sites) is attractive, but incorrectIf failed link may come back up, without a disconnected site ever being aware that it was disconnectedThe site then has old values, and a read from that site would return an incorrect valueIf site was aware of failure reintegration could have been performed, but no way to guarantee thisWith network partitioning, sites in each partition may update same item concurrentlybelieving sites in other partitions have all failed

اسلاید 57: Site ReintegrationWhen failed site recovers, it must catch up with all updates that it missed while it was downProblem: updates may be happening to items whose replica is stored at the site while the site is recoveringSolution 1: halt all updates on system while reintegrating a siteUnacceptable disruptionSolution 2: lock all replicas of all data items at the site, update to latest version, then release locksOther solutions with better concurrency also available

اسلاید 58: Comparison with Remote BackupRemote backup (hot spare) systems (Section 17.10) are also designed to provide high availability Remote backup systems are simpler and have lower overheadAll actions performed at a single site, and only log records shippedNo need for distributed concurrency control, or 2 phase commitUsing distributed databases with replicas of data items can provide higher availability by having multiple (> 2) replicas and using the majority protocolAlso avoid failure detection and switchover time associated with remote backup systems

اسلاید 59: Coordinator SelectionBackup coordinatorssite which maintains enough information locally to assume the role of coordinator if the actual coordinator fails executes the same algorithms and maintains the same internal state information as the actual coordinator fails executes state information as the actual coordinator allows fast recovery from coordinator failure but involves overhead during normal processing.Election algorithmsused to elect a new coordinator in case of failures Example: Bully Algorithm - applicable to systems where every site can send a message to every other site.

اسلاید 60: Bully AlgorithmIf site Si sends a request that is not answered by the coordinator within a time interval T, assume that the coordinator has failed Si tries to elect itself as the new coordinator.Si sends an election message to every site with a higher identification number, Si then waits for any of these processes to answer within T.If no response within T, assume that all sites with number greater than i have failed, Si elects itself the new coordinator.If answer is received Si begins time interval T’, waiting to receive a message that a site with a higher identification number has been elected.

اسلاید 61: Bully Algorithm (Cont.)If no message is sent within T’, assume the site with a higher number has failed; Si restarts the algorithm.After a failed site recovers, it immediately begins execution of the same algorithm.If there are no active sites with higher numbers, the recovered site forces all processes with lower numbers to let it become the coordinator site, even if there is a currently active coordinator with a lower number.

اسلاید 62: Distributed Query ProcessingFor centralized systems, the primary criterion for measuring the cost of a particular strategy is the number of disk accesses.In a distributed system, other issues must be taken into account:The cost of a data transmission over the network.The potential gain in performance from having several sites process parts of the query in parallel.

اسلاید 63: Query TransformationTranslating algebraic queries on fragments.It must be possible to construct relation r from its fragmentsReplace relation r by the expression to construct relation r from its fragmentsConsider the horizontal fragmentation of the account relation intoaccount1 =  branch_name = “Hillside” (account )account2 =  branch_name = “Valleyview” (account )The query  branch_name = “Hillside” (account ) becomes branch_name = “Hillside” (account1  account2)which is optimized into branch_name = “Hillside” (account1)   branch_name = “Hillside” (account2)

اسلاید 64: Example Query (Cont.)Since account1 has only tuples pertaining to the Hillside branch, we can eliminate the selection operation.Apply the definition of account2 to obtain branch_name = “Hillside” ( branch_name = “Valleyview” (account )This expression is the empty set regardless of the contents of the account relation.Final strategy is for the Hillside site to return account1 as the result of the query.

اسلاید 65: Simple Join ProcessingConsider the following relational algebra expression in which the three relations are neither replicated nor fragmentedaccount depositor branchaccount is stored at site S1depositor at S2branch at S3For a query issued at site SI, the system needs to produce the result at site SI

اسلاید 66: Possible Query Processing StrategiesShip copies of all three relations to site SI and choose a strategy for processing the entire locally at site SI.Ship a copy of the account relation to site S2 and compute temp1 = account depositor at S2. Ship temp1 from S2 to S3, and compute temp2 = temp1 branch at S3. Ship the result temp2 to SI.Devise similar strategies, exchanging the roles S1, S2, S3Must consider following factors:amount of data being shipped cost of transmitting a data block between sitesrelative processing speed at each site

اسلاید 67: Semijoin StrategyLet r1 be a relation with schema R1 stores at site S1Let r2 be a relation with schema R2 stores at site S2Evaluate the expression r1 r2 and obtain the result at S1.1. Compute temp1  R1  R2 (r1) at S1.2. Ship temp1 from S1 to S2.3. Compute temp2  r2 temp1 at S24. Ship temp2 from S2 to S1.5. Compute r1 temp2 at S1. This is the same as r1 r2.

اسلاید 68: Formal DefinitionThe semijoin of r1 with r2, is denoted by:r1 r2 it is defined by:R1 (r1 r2) Thus, r1 r2 selects those tuples of r1 that contributed to r1 r2.In step 3 above, temp2=r2 r1.For joins of several relations, the above strategy can be extended to a series of semijoin steps.

اسلاید 69: Join Strategies that Exploit ParallelismConsider r1 r2 r3 r4 where relation ri is stored at site Si. The result must be presented at site S1.r1 is shipped to S2 and r1 r2 is computed at S2: simultaneously r3 is shipped to S4 and r3 r4 is computed at S4S2 ships tuples of (r1 r2) to S1 as they produced; S4 ships tuples of (r3 r4) to S1 Once tuples of (r1 r2) and (r3 r4) arrive at S1 (r1 r2) (r3 r4) is computed in parallel with the computation of (r1 r2) at S2 and the computation of (r3 r4) at S4.

اسلاید 70: Heterogeneous Distributed DatabasesMany database applications require data from a variety of preexisting databases located in a heterogeneous collection of hardware and software platformsData models may differ (hierarchical, relational , etc.)Transaction commit protocols may be incompatibleConcurrency control may be based on different techniques (locking, timestamping, etc.)System-level details almost certainly are totally incompatible.A multidatabase system is a software layer on top of existing database systems, which is designed to manipulate information in heterogeneous databasesCreates an illusion of logical database integration without any physical database integration

اسلاید 71: AdvantagesPreservation of investment in existinghardwaresystem softwareApplicationsLocal autonomy and administrative control Allows use of special-purpose DBMSsStep towards a unified homogeneous DBMSFull integration into a homogeneous DBMS facesTechnical difficulties and cost of conversionOrganizational/political difficultiesOrganizations do not want to give up control on their dataLocal databases wish to retain a great deal of autonomy

اسلاید 72: Unified View of DataAgreement on a common data modelTypically the relational modelAgreement on a common conceptual schemaDifferent names for same relation/attributeSame relation/attribute name means different thingsAgreement on a single representation of shared data E.g. data types, precision, Character setsASCII vs EBCDICSort order variationsAgreement on units of measure Variations in namesE.g. Köln vs Cologne, Mumbai vs Bombay

اسلاید 73: Query ProcessingSeveral issues in query processing in a heterogeneous databaseSchema translationWrite a wrapper for each data source to translate data to a global schemaWrappers must also translate updates on global schema to updates on local schemaLimited query capabilitiesSome data sources allow only restricted forms of selectionsE.g. web forms, flat file data sourcesQueries have to be broken up and processed partly at the source and partly at a different siteRemoval of duplicate information when sites have overlapping informationDecide which sites to execute queryGlobal query optimization

اسلاید 74: Mediator SystemsMediator systems are systems that integrate multiple heterogeneous data sources by providing an integrated global view, and providing query facilities on global viewUnlike full fledged multidatabase systems, mediators generally do not bother about transaction processingBut the terms mediator and multidatabase are sometimes used interchangeablyThe term virtual database is also used to refer to mediator/multidatabase systems

اسلاید 75: Directory SystemsTypical kinds of directory informationEmployee information such as name, id, email, phone, office addr, ..Even personal information to be accessed from multiple placese.g. Web browser bookmarksWhite pagesEntries organized by name or identifierMeant for forward lookup to find more about an entryYellow pagesEntries organized by propertiesFor reverse lookup to find entries matching specific requirementsWhen directories are to be accessed across an organizationAlternative 1: Web interface. Not great for programsAlternative 2: Specialized directory access protocolsCoupled with specialized user interfaces

اسلاید 76: Directory Access ProtocolsMost commonly used directory access protocol:LDAP (Lightweight Directory Access Protocol)Simplified from earlier X.500 protocolQuestion: Why not use database protocols like ODBC/JDBC?Answer: Simplified protocols for a limited type of data access, evolved parallel to ODBC/JDBCProvide a nice hierarchical naming mechanism similar to file system directoriesData can be partitioned amongst multiple servers for different parts of the hierarchy, yet give a single view to userE.g. different servers for Bell Labs Murray Hill and Bell Labs BangaloreDirectories may use databases as storage mechanism

اسلاید 77: LDAP: Lightweight Directory Access ProtocolLDAP Data ModelData ManipulationDistributed Directory Trees

اسلاید 78: LDAP Data ModelLDAP directories store entriesEntries are similar to objectsEach entry must have unique distinguished name (DN)DN made up of a sequence of relative distinguished names (RDNs)E.g. of a DNcn=Silberschatz, ou-Bell Labs, o=Lucent, c=USAStandard RDNs (can be specified as part of schema)cn: common name ou: organizational unito: organization c: countrySimilar to paths in a file system but written in reverse direction

اسلاید 79: LDAP Data Model (Cont.)Entries can have attributesAttributes are multi-valued by defaultLDAP has several built-in types Binary, string, time typesTel: telephone number PostalAddress: postal addressLDAP allows definition of object classes Object classes specify attribute names and typesCan use inheritance to define object classesEntry can be specified to be of one or more object classesNo need to have single most-specific type

اسلاید 80: LDAP Data Model (cont.)Entries organized into a directory information tree according to their DNsLeaf level usually represent specific objectsInternal node entries represent objects such as organizational units, organizations or countriesChildren of a node inherit the DN of the parent, and add on RDNs E.g. internal node with DN c=USAChildren nodes have DN starting with c=USA and further RDNs such as o or ouDN of an entry can be generated by traversing path from rootLeaf level can be an alias pointing to another entryEntries can thus have more than one DNE.g. person in more than one organizational unit

اسلاید 81: LDAP Data ManipulationUnlike SQL, LDAP does not define DDL or DMLInstead, it defines a network protocol for DDL and DMLUsers use an API or vendor specific front endsLDAP also defines a file format LDAP Data Interchange Format (LDIF) Querying mechanism is very simple: only selection & projection

اسلاید 82: LDAP QueriesLDAP query must specifyBase: a node in the DIT from where search is to startA search conditionBoolean combination of conditions on attributes of entriesEquality, wild-cards and approximate equality supportedA scopeJust the base, the base and its children, or the entire subtree from the baseAttributes to be returnedLimits on number of results and on resource consumptionMay also specify whether to automatically dereference aliasesLDAP URLs are one way of specifying queryLDAP API is another alternative

اسلاید 83: LDAP URLsFirst part of URL specifis server and DN of baseldap:://aura.research.bell-labs.com/o=Lucent,c=USAOptional further parts separated by ? symbolldap:://aura.research.bell-labs.com/o=Lucent,c=USA??sub?cn=KorthOptional parts specifyattributes to return (empty means all)Scope (sub indicates entire subtree)Search condition (cn=Korth)

اسلاید 84: C Code using LDAP API #include <stdio.h> #include <ldap.h> main( ) { LDAP *ld; LDAPMessage *res, *entry; char *dn, *attr, *attrList [ ] = {“telephoneNumber”, NULL}; BerElement *ptr; int vals, i; // Open a connection to server ld = ldap_open(“aura.research.bell-labs.com”, LDAP_PORT); ldap_simple_bind(ld, “avi”, “avi-passwd”); … actual query (next slide) … ldap_unbind(ld); }

اسلاید 85: C Code using LDAP API (Cont.)ldap_search_s(ld, “o=Lucent, c=USA”, LDAP_SCOPE_SUBTREE, “cn=Korth”, attrList, /* attrsonly*/ 0, &res); /*attrsonly = 1 => return only schema not actual results*/ printf(“found%d entries”, ldap_count_entries(ld, res)); for (entry=ldap_first_entry(ld, res); entry != NULL; entry=ldap_next_entry(id, entry)) { dn = ldap_get_dn(ld, entry); printf(“dn: %s”, dn); /* dn: DN of matching entry */ ldap_memfree(dn); for(attr = ldap_first_attribute(ld, entry, &ptr); attr != NULL; attr = ldap_next_attribute(ld, entry, ptr)) { // for each attribute printf(“%s:”, attr); // print name of attribute vals = ldap_get_values(ld, entry, attr); for (i = 0; vals[i] != NULL; i ++) printf(“%s”, vals[i]); // since attrs can be multivalued ldap_value_free(vals); } } ldap_msgfree(res);

اسلاید 86: LDAP API (Cont.)LDAP API also has functions to create, update and delete entriesEach function call behaves as a separate transactionLDAP does not support atomicity of updates

اسلاید 87: Distributed Directory TreesOrganizational information may be split into multiple directory information treesSuffix of a DIT gives RDN to be tagged onto to all entries to get an overall DNE.g. two DITs, one with suffix o=Lucent, c=USA and another with suffix o=Lucent, c=IndiaOrganizations often split up DITs based on geographical location or by organizational structureMany LDAP implementations support replication (master-slave or multi-master replication) of DITs (not part of LDAP 3 standard)A node in a DIT may be a referral to a node in another DITE.g. Ou= Bell Labs may have a separate DIT, and DIT for o=Lucent may have a leaf with ou=Bell Labs containing a referral to the Bell Labs DITReferalls are the key to integrating a distributed collection of directoriesWhen a server gets a query reaching a referral node, it may eitherForward query to referred DIT and return answer to client, orGive referral back to client, which transparently sends query to referred DIT (without user intervention)

اسلاید 88: End of Chapter

اسلاید 89: Three Phase Commit (3PC)Assumptions:No network partitioningAt any point, at least one site must be up.At most K sites (participants as well as coordinator) can failPhase 1: Obtaining Preliminary Decision: Identical to 2PC Phase 1.Every site is ready to commit if instructed to do soPhase 2 of 2PC is split into 2 phases, Phase 2 and Phase 3 of 3PCIn phase 2 coordinator makes a decision as in 2PC (called the pre-commit decision) and records it in multiple (at least K) sitesIn phase 3, coordinator sends commit/abort message to all participating sites,Under 3PC, knowledge of pre-commit decision can be used to commit despite coordinator failure Avoids blocking problem as long as < K sites failDrawbacks: higher overheadsassumptions may not be satisfied in practice

اسلاید 90: Figure 22.3

اسلاید 91: Figure 22.4

اسلاید 92: Figure 22.5

اسلاید 93: Figure 22.7