DistributedDatabasesPakistan

DistributedDatabasesPakistan - Distributed Databases...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Distributed Databases Distributed Dr. Julian Bunn Center for Advanced Computing Research Caltech Based on material provided by: Jim Gray (Microsoft), Heinz Stockinger (CERN), Raghu Ramakrishnan (Wisconsin) Outline ? ? ? ? Introduction to Database Systems Distributed Databases Distributed Systems Distributed Databases for Physics J.J.Bunn, Distributed Databases, 2001 2 Part I Part Introduction to Database Systems . Julian Bunn California Institute of Technology What is a Database? ? ? ? ? ? A large, integrated collection of data Entities (things) and Relationships (connections) Objects and Associations/References A Database Management System (DBMS) is a software package designed to store and manage Databases “Traditional” (ER) Databases and “Object” Databases J.J.Bunn, Distributed Databases, 2001 4 Why Use a DBMS? ? ? ? ? ? ? ? ? ? ? Data Independence Efficient Access Reduced Application Development Time Data Integrity Data Security Data Analysis Tools Uniform Data Administration Concurrent Access Automatic Parallelism Recovery from crashes 5 J.J.Bunn, Distributed Databases, 2001 Cutting Edge Databases ? ? ? ? ? Scientific Applications Digital Libraries, Interactive Video, Human Genome project, Particle Physics Experiments, National Digital Observatories, Earth Images Commercial Web Systems Data Mining / Data Warehouse Simple data but very high transaction rate and enormous volume (e.g. click through) 6 J.J.Bunn, Distributed Databases, 2001 Data Models ? ? ? Data Model: A Collection of Concepts for Describing Data Schema: A Set of Descriptions of a Particular Collection of Data, in the context of the Data Model Relational Model: ? E.g. A Lecture is attended by zero or more is Students E.g. A Database Lecture inherits attributes inherits from a general Lecture 7 ? Object Model: ? J.J.Bunn, Distributed Databases, 2001 Data Independence ? Applications insulated from how data in the Database is structured and stored Logical Data Independence: Protection from changes in the logical structure of the data ? Physical Data Independence: Protection from changes in the physical structure of the data ? J.J.Bunn, Distributed Databases, 2001 8 Concurrency Control ? ? Good DBMS performance relies on allowing concurrent access to the data by more than one client DBMS ensures that interleaved actions coming from different clients do not cause inconsistency in the data ? E.g. two simultaneous bookings for the same airplane seat ? Each client is unaware of how many other clients are using the DBMS 9 J.J.Bunn, Distributed Databases, 2001 Transactions ? ? A Transaction is an atomic sequence of actions in the Database (reads and writes) Each Transaction has to be executed completely, and must leave the completely Database in a consistent state ? The definition of “consistent” is ultimately the client’s responsibility! responsibility! ? If the Transaction fails or aborts midway, then the Database is “rolled back” to its initial consistent state (when the Transaction began). 10 J.J.Bunn, Distributed Databases, 2001 What Is A Transaction? ? ? Programmer’s view: ? Bracket a collection of actions Only two outcomes: Begin() action action action A simple failure model simple ? Begin() action action action action Begin() action action action Rollback() Rollback() Commit() Fail !! Fail Success! J.J.Bunn, Distributed Databases, 2001 Failure! 11 ACID ? ? ? ? Atomic: all or nothing Consistent: state transformation Isolated: no concurrency anomalies Durable: committed transaction effects persist J.J.Bunn, Distributed Databases, 2001 12 Why Bother: Atomicity? ? RPC semantics: ? ? At most once: try one time At least once: keep trying ’till acknowledged Exactly once: keep trying ’till acknowledged and server discards duplicate requests ? ? ? ? J.J.Bunn, Distributed Databases, 2001 13 Why Bother: Atomicity? ? Example: insert record in file ? At most once: time-out means “maybe” ? At least once: retry may get “duplicate” error or retry may do second insert ? Exactly once: you do not have to worry ? What if operation involves ? Insert several records? ? Send several messages? ? Want ALL or NOTHING for group of actions J.J.Bunn, Distributed Databases, 2001 14 Why Bother: Consistency ? ? Begin-Commit brackets a set of operations You can violate consistency inside brackets Debit but not credit (destroys money) ? Delete old file before create new file in a copy ? Print document before delete from spool queue ? ? Begin and commit are points of consistency State transformations new state under construction Commit Begin J.J.Bunn, Distributed Databases, 2001 15 Why Bother: Isolation ? Running programs concurrently on same data can create concurrency anomalies ? The shared checking account example Begin() read BAL add 10 write BAL Bal = 100 Bal = 100 Bal = 110 Bal = 70 Begin() read BAL Subtract 30 write BAL Commit() Commit() ? Programming is hard enough without having to worry about concurrency 16 J.J.Bunn, Distributed Databases, 2001 Isolation ? ? It is as though programs run one at a time ? No concurrency anomalies System automatically protects applications ? Locking (DB2, Informix, Microsoft® SQL Server™, Sybase…) SQL ? Versioned databases (Oracle, Interbase…) Begin() read BAL add 10 write BAL Bal = 100 Begin() Bal = 110 Bal = 110 Bal = 80 Commit() read BAL Subtract 30 write BAL Commit() J.J.Bunn, Distributed Databases, 2001 17 Why Bother: Durability ? ? Once a transaction commits, want effects to survive failures Fault tolerance: old master-new master won’t work: Can’t do daily dumps: would lose recent work ? Want “continuous” dumps ? ? ? Redo “lost” transactions in case of failure Resend unacknowledged messages 18 J.J.Bunn, Distributed Databases, 2001 Why ACID For Client/Server And Distributed ? ? ? ACID is important for centralized systems Failures in centralized systems are simpler In distributed systems: ? More and more -independent failures ? ACID is harder to implement ? That makes it even MORE IMPORTANT ? Simple failure model ? Simple repair model J.J.Bunn, Distributed Databases, 2001 19 ACID Generalizations ? Taxonomy of actions ? Unprotected: ? Temp not undone or redone files ? Transactional: can be undone before commit ? Database and message operations ? Real: cannot be undone ? Drill a hole in a piece of metal, print a check ? ? Nested transactions: subtransactions Work flow: long-lived transactions 20 J.J.Bunn, Distributed Databases, 2001 Scheduling Transactions ? ? ? ? The DBMS has to take care of a set of Transactions that arrive concurrently It converts the concurrent Transaction set into a new set that can be executed sequentially It ensures that, before reading or writing an Object, each Transaction waits for a Lock on the Object Lock Each Transaction releases all its Locks when finished ? (Strict Two -Phase-Locking Protocol) Two-Phase21 J.J.Bunn, Distributed Databases, 2001 Concurrency Control Locking ? ? How to automatically prevent concurrency bugs? Serialization theorem: ? ? If you lock all you touch and hold to commit: no bugs If you do not follow these rules, you may see bugs Set automatically (well-formed) Released at commit/rollback (two-phase locking) Granularity: objects or containers or server Mode: shared or exclusive or… 22 ? Automatic Locking: ? ? ? Greater concurrency for locks: ? ? J.J.Bunn, Distributed Databases, 2001 Reduced Isolation Levels ? ? ? It is possible to lock less and risk fuzzy data Example: want statistical summary of DB ? But do not want to lock whole database Repeatable Read: may see fuzzy inserts/delete ? But will serialize all updates Read Committed: see only committed data Read Uncommitted: may see uncommitted updates Reduced levels: ? ? ? J.J.Bunn, Distributed Databases, 2001 23 Ensuring Atomicity ? ? ? The DBMS ensures the atomicity of a atomicity Transaction, even if the system crashes in the middle of it In other words all of the Transaction is all applied to the Database, or none of it is none How? ? ? ? Keep a log/history of all actions carried out on log/history the Database Before making a change, put the log for the change somewhere “safe” After a crash, effects of partially executed transactions are undone using the log 24 J.J.Bunn, Distributed Databases, 2001 DO/UNDO/REDO ? Each action generates a log record Old state DO New state ? Has an UNDO action Log New New state UNDO Old state Log ? Has a REDO action Old state Log New New state REDO J.J.Bunn, Distributed Databases, 2001 25 What Does A Log Record Look Like? ? Log record has ? Header ? Item (transaction ID, timestamp… ) ID ? Old value ? New value ? ? ? ? Log ? Log For messages: just message text and sequence # For records: old and new value on update Keep records small 26 J.J.Bunn, Distributed Databases, 2001 Transaction Is A Sequence Of Actions ? Each action changes state ? ? ? Changes database Sends messages Operates a display/printer/drill press Old state Old state Old state Old state DO Log Log DO New state New DO New state New DO New state New New state Log Log Log ? Leaves a log trail J.J.Bunn, Distributed Databases, 2001 27 Transaction UNDO Is Easy ? ? ? Read log backwards UNDO one step at a time Can go half-way back to get nested transactions Old state Old Old state New state New state UNDO New state Old Old state UNDO New Old state UNDO New state Log UNDO Log Log Log Log J.J.Bunn, Distributed Databases, 2001 28 Durability: Protecting The Log ? When transaction commits ? Put its log in a durable place (duplexed disk) ? Need log to redo transaction in case of failure ? System failure: lost in-memory updates Log LLog og Log Log LLog og Log Log Log Log Log Log LLog og ? Media failure (lost disk) ? ? This makes transaction durable Log is sequential file ? Converts random IO to single sequential IO ? See NTFS or newer UNIX file systems J.J.Bunn, Distributed Databases, 2001 W rit e 29 Recovery After System Failure ? ? During normal processing, write checkpoints on non -volatile storage When recovering from a system failure… ? return to the checkpoint state ? Reapply log of all committed transactions ? Force-at-commit insures log will survive restart ? Then UNDO all uncommitted transactions REDO Old state REDO New state New state Old Old state REDO REDO Log LogREDO Log Log Log Old state Old state New state New New state J.J.Bunn, Distributed Databases, 2001 30 Dealing with failure ? ? Idempotence What if fail during restart? ? REDO many times UNDO something not done What if new state not around at restart? ? Old state REDO Log New state REDO REDO Log Log New state New state UNDO Log Old state UNDO UNDO Log Log Old state J.J.Bunn, Distributed Databases, 2001 31 Dealing with failure ? Idempotence Solution: make F(F(x))=F(x) (idempotence) ? ? Discard duplicates ? Message sequence numbers to discard duplicates ? Use sequence numbers on pages to detect state (Or) make operations idempotent ? Move to position x, write value V to byte B… New state New state REDO REDO Log Log Log New state UNDO Log Log Old state UNDO UNDO Old state Old state REDO Log J.J.Bunn, Distributed Databases, 2001 32 The Log: More Detail ? Actions recorded in the Log Transaction writes an Object ? Store in the Log: Transaction Identifier, Object Identifier, new value and old value ? This must happen before actually writing the Object! ? Transaction commits or aborts ? ? ? Duplicate Log on “stable” storage Log records chained by Transaction Identifier: easy to undo a Transaction 33 J.J.Bunn, Distributed Databases, 2001 Structure of a Database ? Typical DBMS has a layered architecture Query Optimisation & Execution Relational Operators Files and Access Methods Buffer Management Disk Space Management Disk J.J.Bunn, Distributed Databases, 2001 34 Database Administration ? ? ? ? Design Logical/Physical Schema Handle Security and Authentication Ensure Data Availability, Crash Recovery Tune Database as needs and workload evolves J.J.Bunn, Distributed Databases, 2001 35 Summary ? ? ? ? ? Databases are used to maintain and query large datasets DBMS benefits include recovery from crashes, concurrent access, data integrity and security, quick application development Abstraction ensures independence ACID Increasingly Important (and Big) in Scientific and Commercial Enterprises 36 J.J.Bunn, Distributed Databases, 2001 Part 2 Part Distributed Databases . Julian Bunn California Institute of Technology Distributed Databases ? Data are stored at several locations ? Each managed by a DBMS that can run autonomously ? Ideally, location of data is unknown to client ? Distributed Data Independence ? Distributed Transactions are supported Clients can write Transactions regardless of where the affected data are located ? Distributed Transaction Atomicity ? Hard, and in some cases undesirable ? ? E.g. need to avoid overhead of ensuring location transparency 38 J.J.Bunn, Distributed Databases, 2001 Types of Distributed Database ? ? Homogeneous: Every site runs the same type of DBMS Heterogeneous: Different sites run different DBMS (maybe even RDBMS and ODBMS) J.J.Bunn, Distributed Databases, 2001 39 ? Client-Servers ? Distributed DBMS Architectures Client sends query to each database server in the distributed system ? Client caches and accumulates responses ? Collaborating Server Client sends query to “nearest” Server ? Server executes query locally ? Server sends query to other Servers, as required ? Server sends response to Client ? J.J.Bunn, Distributed Databases, 2001 40 Storing the Distributed Data ? In fragments at each site Split the data up ? Each site stores one or more fragments ? ? In complete replicas at each site ? Each site stores a replica of the complete data Each site stores some replicas and/or fragments or the data 41 ? A mixture of fragments and replicas ? J.J.Bunn, Distributed Databases, 2001 Partitioned Data Break file into disjoint groups ? Exploit data access locality ? ? ? ? ? Orders N.A. S.A. Europe Asia Put data near consumer Less network traffic Better response time Better availability Owner controls data autonomy data or traffic may exceed single store 42 ? Spread Load ? J.J.Bunn, Distributed Databases, 2001 How to Partition Data? ? How to Partition ? ? ? ? by attribute or random or by source or by use Directory (replicated) or Algorithm N.A. S.A. Europe Asia ? Problem: to find it must have ? ? ? Encourages attribute -based partitioning J.J.Bunn, Distributed Databases, 2001 43 Replicated Data Place fragment at many sites ? Pros: + Improves availability + Disconnected (mobile) operation + Distributes load + Reads are cheaper Cons: ? N times more updates ? N times more storage Placement strategies: ? Dynamic: cache on demand ? Static: place specific Catalog ? ? J.J.Bunn, Distributed Databases, 2001 44 Fragmentation ? ? Horizontal – “Row-wise” ? E.g. rows of the table make up one fragment E.g. columns of the table make up one fragment Vertical – “Column-Wise” ? ID #Particles … … 10001 3 10002 3 10003 4 10004 5 10005 6 10006 6 10007 6 10008 9 … … J.J.Bunn, Distributed Databases, 2001 Energy … 121.5 202.2 99.3 231.9 287.1 107.7 98.9 100.1 … Event# … 111 112 113 120 125 126 127 128 … Run# … 13120 13120 13120 13120 13120 13120 13120 13120 … Date … 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 … Time … 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 … 45 Replication ? Make synchronised or unsynchronised copies of data at servers Synchronised : data are always current, updates are constantly shipped between replicas ? Unsynchronised: good for read-only data ? ? ? Increases availability of data Makes query execution faster J.J.Bunn, Distributed Databases, 2001 46 Distributed Catalogue Management ? ? Need to know where data are distributed in the system At each site, need to name each replica of each data fragment ? “Local name”, “Birth Place” Describes all fragments and replicas at the site Keeps track of replicas of relations at the site To find a relation, look up Birth site’s catalogue: “Birth Place” site never changes, even if relation is moved 47 ? Site Catalogue: ? ? ? J.J.Bunn, Distributed Databases, 2001 Replication Catalogue ? ? ? ? Which objects are being replicated Where objects are being replicated to How updates are propagated Catalogue is a set of tables that can be backed up, and recovered (as any other table) These tables are themselves replicated to each replication site ? ? No single point of failure in the Distributed Database 48 J.J.Bunn, Distributed Databases, 2001 Configurations ? ? ? ? ? Single Master with multiple read-only snapshot sites Multiple Masters Single Master with multiple updatable snapshot sites Master at record-level granularity Hybrids of the above J.J.Bunn, Distributed Databases, 2001 49 Distributed Queries Islamabad ID #Particles … … 10001 3 10002 3 10003 4 10004 5 10005 6 10006 6 10007 6 10008 9 … … Energy … 121.5 202.2 99.3 231.9 287.1 107.7 98.9 100.1 … Event# … 111 112 113 120 125 126 127 128 … Run# … 13120 13120 13120 13120 13120 13120 13120 13120 … Date … 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 … Time … 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 … ID #Particles … … 10001 3 10002 3 10003 4 10004 5 10005 6 10006 6 10007 6 10008 9 … … Geneva Energy … 121.5 202.2 99.3 231.9 287.1 107.7 98.9 100.1 … Event# … 111 112 113 120 125 126 127 128 … Run# … 13120 13120 13120 13120 13120 13120 13120 13120 … Date … 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 … Time … 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 … ? ? ? SELECT AVG(E.Energy) FROM Events E WHERE E.particles > 3 AND E.particles < 7 Replicated: Copies of the complete Event table at Geneva and at Islamabad Choice of where to execute query ? Based on local costs, network costs, remote capacity, etc. 50 J.J.Bunn, Distributed Databases, 2001 Distributed Queries (contd.) ? SELECT AVG(E.Energy) FROM Events E WHERE E.particles > 3 AND E.particles < 7 Row-wise fragmented: ID #Particles … … 10001 3 10002 3 10003 4 10004 5 10005 6 10006 6 10007 6 10008 9 … … Energy … 121.5 202.2 99.3 231.9 287.1 107.7 98.9 100.1 … Event# … 111 112 113 120 125 126 127 128 … Run# … 13120 13120 13120 13120 13120 13120 13120 13120 … ? Date … 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 … Time … 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 … Particles < 5 at Geneva, Particles > 4 at Islamabad ? ? Need to compute SUM(E.Energy) and COUNT(E.Energy) at both sites both If WHERE clause had E.particles > 4 then only need to compute at Islamabad 51 J.J.Bunn, Distributed Databases, 2001 Distributed Queries (contd.) ? SELECT AVG(E.Energy) FROM Events E WHERE E.particles > 3 AND E.particles < 7 ID #Particles … … 10001 3 10002 3 10003 4 10004 5 10005 6 10006 6 10007 6 10008 9 … … Energy … 121.5 202.2 99.3 231.9 287.1 107.7 98.9 100.1 … Event# … 111 112 113 120 125 126 127 128 … Run# … 13120 13120 13120 13120 13120 13120 13120 13120 … Date … 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 3/1406 … Time … 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 13:30:55.0001 … ? Column-wise Fragmented: ID, Energy and Event# Columns at Geneva, ID and remaining Columns at Islamabad: ? ? ? Need to join on ID Select IDs satisfying Particles constraint at Islamabad SUM(Energy) and Count(Energy) for those IDs at Geneva J.J.Bunn, Distributed Databases, 2001 52 Joins ? ? ? ? Joins are used to compare or combine relations (rows) from two or more tables, when the relations share a common attribute value Simple approach: for every relation in the first table “S”, loop over all relations in the other table “R”, and see if the attributes match N-way joins are evaluated as a series of 2-way joins Join Algorithms are a continuing topic of intense research in Computer Science 53 J.J.Bunn, Distributed Databases, 2001 Join Algorithms ? ? ? ? ? ? Need to run in memory for best performance Nested-Loops: efficient only if “R” very small (can be stored in memory) Hash-Join: Build an in-memory hash table of “R”, then loop over “S” hashing to check for match Hybrid Hash-Join: When “R” hash is too big to fit in memory, split join into partitions Merge-Join: Used when “R” and “S” are already sorted on the join attribute, simply merging them in parallel Special versions of Join Algorithms needed for Distributed Database query execution! 54 J.J.Bunn, Distributed Databases, 2001 Distributed Query Optimisation ? Cost-based: Consider all “plans” ? Pick cheapest: include communication costs ? ? ? Need to use distributed join methods Site that receives query constructs Global Plan, hints for local plans ? Local plans may be changed at each site 55 J.J.Bunn, Distributed Databases, 2001 Replication ? ? Synchronous: All data that have been changed must be propagated before the Transaction commits Asynchronous: Changed data are periodically sent Replicas may go out of sync. ? Clients must be aware of this ? J.J.Bunn, Distributed Databases, 2001 56 Synchronous Replication Costs ? Before an update Transaction can commit, it obtains locks on all modified copies Sends lock requests to remote sites, holds locks ? If links or remote sites fail, Transaction cannot commit until links/sites restored ? Even without failure, commit protocol is commit complex, and involves many messages ? J.J.Bunn, Distributed Databases, 2001 57 Asynchronous Replication ? ? Allows Transaction to commit before all copies have been modified Two methods: ? Primary Site Peer-to-Peer ? J.J.Bunn, Distributed Databases, 2001 58 Primary Site Replication One copy designated as “Master” ? Published to other sites who subscribe to “Secondary” copies ? Changes propagated to “Secondary” copies ? Done in two steps: ? Capture changes made by committed Transactions ? Apply these changes ? J.J.Bunn, Distributed Databases, 2001 59 The Capture Step ? Procedural: A procedure, automatically invoked, does the capture (takes a snapshot) Log-based: the log is used to generate a Change Data Table ? ? Better (cheaper and faster) but relies on proprietary log details J.J.Bunn, Distributed Databases, 2001 60 The Apply Step ? The Secondary site periodically obtains from the Primary site a snapshot or changes to the Change Data Table Updates its copy ? Period can be timer-based or defined by the user/application ? ? Log-based capture with continuous Apply minimises delays in propagating changes 61 J.J.Bunn, Distributed Databases, 2001 Peer to P Replication - - eer ? ? ? ? More than one copy can be “Master” Changes are somehow propagated to other copies Conflicting changes must be resolved So best when conflicts do not or cannot arise: Each “Master” owns a disjoint fragment or copy ? Update permission only granted to one “Master” at a time ? J.J.Bunn, Distributed Databases, 2001 62 Replication Examples ? Master copy, many slave copies (SQL Server) ? ? always know the correct value (master) change propagation can be ? transactional ? as soon as possible ? periodic ? on demand allows mobile (disconnected) updates updates propagated ASAP, periodic, on demand non-serializable colliding updates must be reconciled. hard to know “real” value 63 ? Symmetric, and anytime (Access) ? ? ? ? ? J.J.Bunn, Distributed Databases, 2001 Data Warehousing and Replication ? Build giant “warehouses” of data from many sites ? Enable complex decision support queries over data from across an organisation ? Warehouses can be seen as an instance of asynchronous replication ? Source data is typically controlled by different DBMS: emphasis on “cleaning” data by removing mismatches while creating replicas ? Procedural Capture and application Apply work best for this environment 64 J.J.Bunn, Distributed Databases, 2001 Distributed Locking ? How to manage Locks across many sites? Centrally: one site does all locking ? Vulnerable to single site failure ? Primary Copy: all locking for an object done at the primary copy site for the object ? Reading requires access to locking site as well as site which stores object ? Fully Distributed: locking for a copy done at site where the copy is stored ? Locks at all sites while writing an object ? J.J.Bunn, Distributed Databases, 2001 65 Distributed Deadlock Detection ? ? Each site maintains a local “waits -for” graph Global deadlock might occur even if local graphs contain no cycles ? ? E.g. Site A holds lock on X, waits for lock on Y Site B holds lock on Y, waits for lock on X Centralised (send all local graphs to one site) Hierarchical (organise sites into hierarchy and send local graphs to parent) Timeout (abort Transaction if it waits too long) 66 ? Three solutions: ? ? ? J.J.Bunn, Distributed Databases, 2001 Distributed Recovery ? ? ? ? Links and Remote Sites may crash/fail If sub-transactions of a Transaction execute at different sites, all or none must commit Need a commit protocol to achieve this Solution: Maintain a Log at each site of commit protocol actions ? Two-Phase Commit 67 J.J.Bunn, Distributed Databases, 2001 Two P - hase Commit ? ? Site which originates Transaction is coordinator, other sites involved in Transaction are subordinates When the Transaction needs to Commit: ? ? ? ? ? ? Coordinator sends “prepare” message to subordinates Subordinates each force-writes an abort or prepare Log abort or prepare record, and sends “yes” or “no” message to Coordinator If Coordinator gets unanimous “yes” messages, force-writes a commit Log record, and sends “commit” message to all commit subordinates Otherwise, force-writes an abort Log record, and sends abort “abort” message to all subordinates Subordinates force-write abort/commit Log record accordingly, then send an “ack” message to Coordinator Coordinator writes end Log record after receiving all acks end J.J.Bunn, Distributed Databases, 2001 68 Notes on Two P - hase Commit (2PC) ? ? ? ? First: voting, Second: termination – both initiated by Coordinator Any site can decide to abort the Transaction Every message is recorded in the local Log by the sender to ensure it survives failures All Commit Protocol log records for a Transaction contain the Transaction ID and Coordinator ID. The Coordinator’s abort/commit record also includes the Site IDs of all subordinates 69 J.J.Bunn, Distributed Databases, 2001 Restart after Site Failure ? If there is a commit or abort Log record for Transaction T, but no end record, then must undo/redo T ? If the site is Coordinator for T, then keep sending commit/abort messages to Subordinates until commit/abort acks received acks ? If there is a prepare Log record, but no commit or abort: ? ? This site is a Subordinate for T Contact Coordinator to find status of T, then ? write commit/abort Log record commit/abort ? Redo/undo T ? Write end Log record end 70 J.J.Bunn, Distributed Databases, 2001 Blocking ? ? ? If Coordinator for Transaction T fails, then Subordinates who have voted “yes” cannot decide whether to commit or abort until Coordinator commit abort until recovers! T is blocked blocked Even if all Subordinates are aware of one another (e.g. via extra information in “prepare” message) they are blocked ? Unless one of them voted “no” 71 J.J.Bunn, Distributed Databases, 2001 Link and Remote Site Failures ? If a Remote Site does not respond during the Commit Protocol for T ? E.g. it crashed or the link is down ? Then If current Site is Coordinator for T: abort ? If Subordinate and not yet voted “yes”: abort ? If Subordinate and has voted “yes”, it is blocked until Coordinator back online ? J.J.Bunn, Distributed Databases, 2001 72 Observations on 2PC ? Ack messages used to let Coordinator know when it can “forget” a Transaction ? Until it receives all acks, it must keep T in the Transaction Table ? ? If Coordinator fails after sending “prepare” messages, but before writing commit/abort Log record, when it comes back up, it aborts T If a subtransaction does no updates, its commit or abort status is irrelevant 73 J.J.Bunn, Distributed Databases, 2001 2PC with Presumed Abort ? When Coordinator aborts T, it undoes T and removes it from the Transaction Table immediately ? ? ? ? ? ? ? Subordinates do not send “ack” on abort If subtransaction does no updates, it responds to “prepare” message with “reader” (instead of “yes”/”no”) Coordinator subsequently ignores “ reader”s If all Subordinates are “reader”s, then 2nd. Phase not required 74 Doesn’t wait for “acks” “Presumes Abort” if T not in Transaction Table Names of Subordinates not recorded in abort abort Log record J.J.Bunn, Distributed Databases, 2001 Replication and Partitioning Compared Base case Scaleup a 1 TPS system to a 2 TPS centralized system ? Central Scaleup 2x more work Partition Scaleup 2x more work Replication Scaleup 4x more work 75 1 TPS server 100 Users 200 Users 2 TPS server Partitioning Two 1 TPS systems Replication Two 2 TPS systems ? 1 TPS server O tps O tps 100 Users 100 Users 2 TPS server 1 tps 1 tps ? 1 TPS server J.J.Bunn, Distributed Databases, 2001 100 Users 100 Users 2 TPS server “Porter” Agent b - ased Distributed Database ? ? Charles Univ, Prague Based on “Aglets” SDK from IBM J.J.Bunn, Distributed Databases, 2001 76 Part 3 Part Distributed Systems . Julian Bunn California Institute of Technology What’s a Distributed System? ? Centralized: everything in one place ? stand-alone PC or Mainframe ? ? Distributed: ? some parts remote ? distributed users ? distributed execution ? distributed data 78 J.J.Bunn, Distributed Databases, 2001 Why Distribute? ? ? No best organization Organisations constantly swing between ? ? Centralized: focus, control, economy Decentralized: adaptive, responsive, competitive ? Why distribute? reflect organisation or application structure ? empower users / producers ? improve service (response / availability) ? distribute load ? use PC technology (economics) ? J.J.Bunn, Distributed Databases, 2001 79 What Should Be Distributed? ? Users and User Interface ? Thin client Trim client Fat client Presentation workflow Business Objects Database ? ? Processing ? Data ? ? Will discuss tradeoffs later 80 J.J.Bunn, Distributed Databases, 2001 Transparency Transparency in Distributed Systems in ? Make distributed system as easy to use and manage as a centralized system Give a Single-System Image ? ? Location transparency: hide fact that object is remote ? hide fact that object has moved ? hide fact that object is partitioned or replicated ? ? J.J.Bunn, Distributed Databases, 2001 Name doesn’t change if object is replicated, partitioned or moved. 81 ? Objects have ? Globally Unique Identifier (GUIDs) ? location(s) = address(es) ? name(s) ? addresses can change ? objects can have many names Naming The basics Address guid ? ? Names are context dependent: ? Jim (Jim @ KGB ?????? ??? ? ? ??? ? ?Jim @ CIA) ? ????? Many naming systems ? ? ? James UNC: \\node\device\dir \dir \dir\object Internet: http://node.domain.root/dir/dir/dir/object LDAP: ldap://ldap.domain.root/o=org,c=US,cn=dir 82 J.J.Bunn, Distributed Databases, 2001 Name Servers in Distributed Systems ? ? ? ? ? Name servers translate names + context to address (+ GUID) Name servers are partitioned (subtrees of name space) Name servers replicate root of name tree Name servers form a hierarchy Distributed data from hell: ? ? ? high read traffic high reliability & availability autonomy 83 J.J.Bunn, Distributed Databases, 2001 Autonomy in Distributed Systems ? ? Owner of site (or node, or application, or database) (or Wants to control it If my part is working, must be able to access & manage it (reorganize, upgrade, add user,…) ? ? Autonomy is ? Essential ? Difficult to implement. ? Conflicts with global consistency examples: naming, authentication, admin… 84 J.J.Bunn, Distributed Databases, 2001 Security The Basics ? Authentication server subject + Authenticator => subject (Yes + token) | No Object ? Security matrix: subject ? who can do what to whom ? Access control list is column of matrix ? “who” is authenticated ID In a distributed system, “who” and “what” and “whom” are distributed objects J.J.Bunn, Distributed Databases, 2001 Permissions ? 85 ? ? ? ? Security Security in Distributed Systems in domain: Security Security domain: nodes with a shared security server. Security domains can have trust relationships: ? A trusts B: A “believes” B when it says this is Jim@B Security domains form a hierarchy. Delegation: passing authority to a server passing when A asks B to do something (e.g. print a file, read a database) when (e.g. B may need A’s authority ? ? J.J.Bunn, Distributed Databases, 2001 Autonomy requires: ? each node is an authenticator ? each node does own security checks Internet Today: ? no trust among domains (fire walls, many passwords) (fire ? trust based on digital signatures 86 The Ideal Distributed System. The ? Clusters ? Cluster is distributed system BUT single location ? manager ? security policy relatively homogeneous ? Clusters use distributed system techniques for ? ? ? communications is ? high bandwidth ? low latency ? low error rate ? ? load distribution ? storage ? execution growth fault tolerance J.J.Bunn, Distributed Databases, 2001 87 Cluster: Shared What? ? Shared Memory Multiprocessor ? ? ? Multiple processors, one memory all devices are local HP V-class an array of nodes all shared common disks VAXcluster + Oracle each device local to a node ownership may change Beowulf,Tandem, SP2, Wolfpack 88 ? Shared Disk Cluster ? ? ? ? Shared Nothing Cluster ? ? ? J.J.Bunn, Distributed Databases, 2001 Distributed Execution Threads and Messages ? Thread is Execution unit (software analog of cpu+memory) threads ? ? Threads execute at a node Threads communicate via Shared memory (local) ? Messages (local and remote) ? shared memory messages J.J.Bunn, Distributed Databases, 2001 89 Peer to P or Client Server - - eer ? Peer-to-Peer is symmetric: ? Either side can send ? Client-server client sends requests ? server sends responses ? simple subset of peer -to-peer ? requ est resp onse J.J.Bunn, Distributed Databases, 2001 90 Connection less or Connected or ? Connection-less ? ? Connected (sessions) (sessions) ?open request contains ? ? ? client id client context work request ? ? ? client authenticated on each message only a single response message e.g. HTTP, NFS v1 - request/reply - close ?client authenticated once ?Messages arrive in order ?Can send many replies (e.g. FTP) (e.g. ? Server has client context (context sensitive) ? e.g. Winsock and ODBC ? HTTP adding connections J.J.Bunn, Distributed Databases, 2001 91 Remote Procedure Call: The key to transparency y = pObj->f(x); x ? Object may be local or remote Methods on object work wherever it is. Local invocation ? f() ? return val; y = Jval;Distributed Databases, 2001 val .J.Bunn, 92 Remote Procedure Call: The The ? key to transparency Remote invocation proxy Obj Local? y = pObj->f(x); x x Gee!! Nice pictures! marshal ? marshal stub x un marshal pObj->f(x) x Obj Local? Obj Local? f() f() return val; un un marshal marshal val marshal val return val; y = val; J.J.Bunn, Distributed Databases, 2001 val val 93 Object Request Broker (ORB) Orchestrates RPC ? ? ? ? ? ? Registers Servers Manages pools of servers Connects clients to servers Does Naming, request-level authorization, Provides transaction coordination (new feature) Old names: ? ? ? Transaction Processing Monitor, Web server, Transaction NetWare J.J.Bunn, Distributed Databases, 2001 Object-Request Broker 94 Using RPC for Transparency ? Partition Transparency Send updates to correct partition x x y = pfile->write(x); x part Local? send send to to correct correct partition partition val J.J.Bunn, Distributed Databases, 2001 un marshal x pObj->write(x) write() return val; marshal val 95 val Using RPC for Transparency ? Replication Transparency Send updates to EACH node x Send Send to to each each replica replica y = pfile->write(x); x J.J.Bunn, Distributed Databases, 2001 val 96 Client/Server Interactions All can be done with RPC ? ? ? Request-Response response may be many messages C C C S S S S S S S 97 Conversational server keeps client context Dispatcher three-tier: complex operation at server ? Queued de-couples client from server allows disconnected operation J.J.Bunn, Distributed Databases, 2001 C Queued Request/Response ? Time-decouples client and server ? Three Transactions Almost real time, ASAP processing Communicate at each other’s convenience Allows mobile (disconnected) operation ? ? ? Disk queues survive client & server failures Submit Perform Response Client J.J.Bunn, Distributed Databases, 2001 Server 98 Why Queued Processing? ? ? Order ? Prioritize requests ambulance dispatcher favors high-priority calls Manage Workflows Build Ship Invoice Pay Deferred processing in mobile apps Interface heterogeneous systems EDI, MOM: Message-Oriented-Middleware DAD: Direct Access to Data ? J.J.Bunn, Distributed Databases, 2001 99 Work Distribution Spectrum ? ? ? ? Presentation and plug-ins Workflow manages session & invokes objects Business objects Database Thin Presentation workflow Fat Business Objects Database J.J.Bunn, Distributed Databases, 2001 Fat Thin 100 Transaction Processing Evolution to Three Tier ? Mainframe Batch processing (centralized) Dumb terminals & Remote Job Entry Intelligent terminals database backends Workflow Systems Object Request Brokers Application Generators Intelligence migrated to clients cards Mainframe ? green screen 3270 Server TP Monitor ? ? ORB Active 101 J.J.Bunn, Distributed Databases, 2001 Web Evolution to Three Tier ? Intelligence migrated to clients (like TP) Web WAIS archie ghopher green screen Character-mode clients, smart servers Server Mosaic ? GUI Browsers - Web file servers GUI Plugins - Web dispatchers - CGI NS & IE ? ? Smart clients - Web dispatcher (ORB) pools of app servers (ISAPI, Viper) workflow scripts at client & server J.J.Bunn, Distributed Databases, 2001 Active 102 PC Evolution to Three Tier ? Intelligence migrated to server Stand-alone PC (centralized) (centralized) ? PC + File & print server message per I/O IO request reply disk I/O ? PC + Database server message per SQL statement message SQL Statement ? ? PC + App server message per transaction message Transaction ActiveX Client, ORB ActiveX server, Xscript 103 J.J.Bunn, Distributed Databases, 2001 The Pattern: Three Tier Computing ? ? ? Clients do presentation, gather input Clients do some workflow (Xscript) Clients send high-level requests to ORB (Object Request Broker) ORB dispatches workflows and business objects -- proxies for client, orchestrate flows & queues Server-side workflow scripts call on distributed business objects to execute task task Presentation workflow ? Business Objects ? Database 104 J.J.Bunn, Distributed Databases, 2001 Web Client HTML VBscritpt JavaScrpt VB or Java Script Engine VB Java plug-ins VB or Java Virt Machine The Three Tiers Middleware Object server Pool ORB ORB TP Monitor Web Server... HTTP+ Internet DCOM Object & Data server . DCOM (oleDB, ODBC,...) 6.2 LU Legacy Gateways 105 IBM J.J.Bunn, Distributed Databases, 2001 ? Manageability ? ? Why Did Everyone Go To Three T - ier? Presentation Business rules must be with data Middleware operations tools Server resources are precious ORB dispatches requests to server pools Put UI processing near user Put shared data processing near shared data workflow ? Performance (scaleability) ? ? ? Technology & Physics ? ? Business Objects Database J.J.Bunn, Distributed Databases, 2001 106 Why Put Business Objects at Server? DAD’sRaw Data MOM’s Business Objects Customer comes to store Takes what he wants Fills out invoice Leaves money for goods Customer comes to store with list Gives list to clerk Clerk gets goods, makes invoice Customer pays clerk, gets goods Easy to build No clerks J.J.Bunn, Distributed Databases, 2001 Easy to manage Clerks controls access Encapsulation 107 Why Server Pools? ? ? Server resources are precious. Clients have 100x more power than server. Pre-allocate everything on server ? ? ? ? preallocate memory pre-open files pre-allocate threads pre-open and authenticate clients N clients x N Servers x F files = N x N x F file opens!!! ? Keep high duty-cycle on objects (re-use them) ? Pool threads, not one per client HTTP Pool of DBC links ? Classic example: TPC-C benchmark ? ? 2 processes IE everything pre-allocated 7,000 clients IIS SQL 108 J.J.Bunn, Distributed Databases, 2001 Classic Mistakes ? ? ? ? Thread per terminal fix: DB server thread pools fix: server pools Process per request (CGI) fix: ISAPI & NSAPI DLLs fix: connection pools Many messages per operation fix: stored procedures fix: server -side objects File open per request fix: cache hot files 109 J.J.Bunn, Distributed Databases, 2001 Distributed Applications need Transactions! ? ? Transactions are key to structuring distributed applications ACID properties ease exception handling Atomic: all or nothing ? Consistent : state transformation ? Isolated: no concurrency anomalies ? Durable : committed transaction effects persist ? J.J.Bunn, Distributed Databases, 2001 110 Programming & Transactions The Application View ? You Start ? ? ? ? (e.g. in TransactSQL): (e.g. Begin [Distributed] Transaction <name> Perform actions Optional Save Transaction <name> Commit or Rollback XID Caller passes you a transaction You return or Rollback. You can Begin / Commit sub -trans. You can use save points Begin Begin RollBack Commit ? You Inherit a XID ? ? ? ? Return RollBack Return 111 J.J.Bunn, Distributed Databases, 2001 Transaction Save Points Backtracking within a transaction ? action action SAVE WORK:5 action SAVE WORK:6 action action SAVE WORK:7 action action ROLLBACK WORK(7) J.J.Bunn, Distributed Databases, 2001 BEGIN WORK:1 action action SAVE WORK:2 action SAVE WORK:3 action action action SAVE WORK:4 action ROLLBACK WORK(2) ? Allows app to cancel parts of a transaction prior to commit This is in most SQL products action action SAVE WORK:8 action COMMIT WORK 112 Chained Transactions ? ? Commit of T1 implicitly begins T2. Carries context forward to next transaction ? ? ? cursors locks other state C o m m i t Transaction #1 Processing context established J.J.Bunn, Distributed Databases, 2001 Transaction #2 B e g i n Processing context used 113 ? ? ? ? ? T1 Going Beyond Flat Transactions Need transactions within transactions Sub-transactions commit only if root does Only root commit is durable. Subtransactions may rollback if so, all its subtransactions rollback Parallel version of nested transactions T12 T121 T11 T111 T113 T13 T131 T132 T122 T123 Nested Transactions T112 T114 T133 114 J.J.Bunn, Distributed Databases, 2001 A Sequence of Transactions Sequence ? ? ? ? Workflow: Application transactions are multi -step ? order, build, ship & invoice, reconcile Presentation Each step is an ACID unit Workflow is a script describing steps Workflow systems Instantiate the scripts ? Drive the scripts ? Allow query against scripts Examples Manufacturing Work In Process (WIP) Queued processing Loan application & approval, Hospital admissions… ? workflow ? Business Objects Database 115 J.J.Bunn, Distributed Databases, 2001 Workflow Scripts ? ? ? ? Workflow scripts are programs (could use VBScript or JavaScript) If step fails, compensation action handles error Events, messages, time, other steps cause step. Workflow controller drives flows Source branch Compensation Action J.J.Bunn, Distributed Databases, 2001 join case loop Step fork 116 Workflow and ACID ? ? ? ? ? Workflow is not Atomic or Isolated Results of a step visible to all Workflow is Consistent and Durable Each flow may take hours, weeks, months Workflow controller keeps flows moving ? maintains context (state) for each flow ? provides a query and operator interface e.g.: “what is the status of Job # 72149?” ? J.J.Bunn, Distributed Databases, 2001 117 ACID Objects Using ACID DBs The easy way to build transactional objects Application uses transactional objects (objects have ACID properties) If object built on top of ACID objects, then object is ACID. ? ? ? SQL Example: New, EnQueue, DeQueue on top of SQL dim c as Customer dim CM as CustomerMgr ... set C = CM.get(CustID) ... C.credit_limit = 1000 ... CM.update(C, CustID) .. 118 ? SQL provides ACID Business Object: Customer Business Object Mgr: CustomerMgr SQL Persistent Programming languages automate this. J.J.Bunn, Distributed Databases, 2001 ACID Objects From Bare Metal The Hard Way to Build Transactional Objects The ? Object Class is a Resource Manager (RM) Resource ? Provides ACID objects from persistent storage ? Provides Undo (on rollback) ? Provides Redo (on restart or media failure) ? Provides Isolation for concurrent ops ? ? ? Microsoft SQL Server, IBM DB2, Oracle,… are Resource managers. Many more coming. RM implementation techniques described later 119 J.J.Bunn, Distributed Databases, 2001 Transaction Manager ? Transaction Manager (TM): manages transaction objects. XID factory ? tracks them ? coordinates them ? n egi b XID TM enlist App call(..XID) ? ? App gets XID from TM Transactional RPC passes XID on all calls ? manages XID inheritance ? RM ? TM manages commit & rollback J.J.Bunn, Distributed Databases, 2001 120 TM Two P - hase Commit Dealing with multiple RMs ? ? ? If all use one RM, then all or none commit If multiple RMs, then need coordination Standard technique: ? Marriage: Do you? I do. I pronounce…Kiss ? Theater: Ready on the set? Ready! Action! Act ? Sailing: Ready about? Ready! Helm’s a-lee! Tack ? Contract law: Escrow agent ? Two-phase commit: ? 1. J.J.Bunn, Distributed Databases, 2001 Voting phase: can you do it? ? 2. If all vote yes, then commit phase: do it! 121 Two P - hase Commit In Pictures ? ? ? ? Transactions managed by TM App gets unique ID (XID) from TM at Begin() XID passed on Transactional RPC RMs Enlist when first do work on XID gin Be XID TM En lis t App J.J.Bunn, Distributed Databases, 2001 Call(..XID..) Call(..XI D..) RM1 En lis t RM2 122 When App Requests Commit Two Phase Commit in Pictures ? ? ? ? TM tracks all RMs enlisted on an XID TM calls enlisted RM’s Prepared() callback If all vote yes, TM calls RM’s Commit() If any vote no, TM calls RM’s Rollback() 4. TM decides Yes, broadcasts 1. Application requests Commit mit 1 om C yes TM 2 2 ee ar ar ep ep Pr Pr 4 it it m m mm Co Co Ye 3 s 5. RMs acknowledge App 6. TM says yes 4 RM1 Yes3 RM2 Ye Ye s s 5 5 123 2. TM broadcasts prepared? J.J.Bunn, Distributed Databases, 2001 3. RMs all vote Yes Implementing Transactions ? Atomicity The DO/UNDO/REDO protocol ? Idempotence ? Two -phase commit ? ? Durability Durable logs ? Force at commit ? ? Isolation ? Locking or versioning 124 J.J.Bunn, Distributed Databases, 2001 Part 4 Part Distributed Databases for Physics . Julian Bunn California Institute of Technology Distributed Databases in Physics ? ? ? Virtual Observatories (e.g. NVO) Gravity Wave Data (e.g. LIGO) Particle Physics (e.g. LHC Experiments) J.J.Bunn, Distributed Databases, 2001 126 Distributed Particle Physics Data ? Next Generation of particle physics experiments are data intensive Acquisition rates of 100 MBytes/second ? At least One PetaByte (1015 Bytes) of raw data per year, per experiment ? Another PetaByte of reconstructed data ? More PetaBytes of simulated data ? Many TeraBytes of MetaData ? ? To be accessed by ~2000 physicists sitting around the globe J.J.Bunn, Distributed Databases, 2001 127 An Ocean of Objects ? ? Access from anywhere to any object in an Ocean of many PetaBytes of objects Approach: Distribute collections of useful objects to where they will be most used ? Move applications to the collection to locations ? Maintain an up-to-date catalogue of collection locations ? Try to balance the global compute resources with the task load from the global clients ? J.J.Bunn, Distributed Databases, 2001 128 RDBMS vs. Object Database •Users send requests into the server queue •all requests must first be serialized through this queue. •to achieve serialization and avoid conflicts, all requests must go through the server queue. •Once through the queue, the server may be able to spawn off multiple threads •DBMS functionality split between the client and server •allowing computing resources to be used •allowing scalability. •clients added without slowing down others, •ODBMS automatically establishes direct, independent, parallel communication paths between clients and servers •servers added to incrementally increase performance without limit. J.J.Bunn, Distributed Databases, 2001 129 Designing the Distributed Database ? ? Problem is: how to handle distributed clients and distributed data whilst maximising client task throughput and use of resources Distributed Databases for: ? ? ? Use middleware that is conscious of the global state of the system: ? ? ? ? ? The physics data The metadata J.J.Bunn, Distributed Databases, 2001 Where are the clients? What data are they asking for? Where are the CPU resources? Where are the Storage resources? How does the global system measure up to it workload, in the past, now and in the future? 130 Distributed Databases for HEP ? Replica synchronisation usually based on small transactions ? ? Replication at the Object level desired ? ? But HEP transactions are large (and long -lived) Objectivity DRO requires dynamic quorum ? bad for unstable WAN links So too difficult – use file replication ? E.g. GDMP Subscription method ? Which Replica to Select? ? Complex decision tree, involving ? Prevailing WAN and Systems conditions ? Objects that the Query “touches” and “needs” ? Where the compute power is ? Where the replicas are ? Existence of previously cached datasets 131 J.J.Bunn, Distributed Databases, 2001 Distributed LHC Databases Today ? ? ? ? Architecture is loosely coupled, autonomous, Object Databases File-based replication with Globus middleware Efficient WAN transport J.J.Bunn, Distributed Databases, 2001 132 ...
View Full Document

This note was uploaded on 12/23/2009 for the course DBST 663 taught by Professor Tba during the Spring '09 term at MD University College.

Ask a homework question - tutors are online