Oracle & sql server comparison 2
- 1. ORACLE & SQL Server Comparison
- 2. Contents
- 3. Performance
- 7. Oracle 11g Top FeaturesPerformance
- 9. Indexing
- 10. Partitioning
- 12. ClusteringConcurrency Model:In multi-user environments, concurrency control ensures that data updates made by one user do not adversely affect those made by other users.
- 13. Main differences of Implementation:Indexing:Indexes are database structures that are created to provide faster path to data. Using indexes can dramatically reduce disk I/O operations, thus increasing the performance of data retrieval.
- 14. The differences in the indexingIndexing(Cont….)
- 15. Indexing(Cont….)
- 16. Partitioning:Partitioning allows large database structures to be decomposed into smaller and more manageable pieces
- 17. The differences partitioningParallel Execution:Parallel execution of SQL operations can vastly improve the performance for operations involving large volumes of data
- 18. The differences in parallel execution
- 19. Oracle will execute INSERT, UPDATE, DELETE, and MERGE statements in parallel
- 20. With SQL Server 2005 INSERT, UPDATE ,DELETE statements are executed serially (MERGE is not supported)Clustering:Clusters are groups of independent servers, or nodes , connected via private network, that work collaboratively as a single system. Clusters allow application s to scale beyond the limits imposed by single node systems when processing loads exceed the capacity of large individual serversPrivate InterconnectNodePublicnetworkPublicnetworkPublicnetworkPublicnetworkClusterwareon each node
- 21. Clustering(Cont..)Only oracle provides real support for clustered configuration: With Real application Clusters , full and transparent can be obtained by simply adding new nodes as the demand increases.
- 22. SQL Server 2000 & 2005 does not support this type of clusteringInformation Integration
- 23. Federation
- 24. Replication
- 25. Message QueuingFederation:Federation provides on-demand access to distributed data, providing a single global view of the data and virtualizing the data’s physical location. Data can be in homogeneous or heterogeneous data stores. Any single piece of data is stored in a single location, and updates and queries are made in real time.
- 27. Supported by both by different manner
- 29. Messaging QueuingMessaging Queuing (Cont)
- 30. Messaging Queuing (Cont)
- 33. System Failures
- 34. Data Failures
- 36. Human Errors
- 38. System Changes
- 39. Data ChangesSystem FailuresReal Application ClustersASMFlashbackRMAN & Oracle Secure BackupData GuardStreams Data FailuresSystem ChangesOnline ReconfigurationRolling UpgradesData ChangesOnline RedefinitionIntegrated High Availability features of Oracle DatabaseUnplannedDowntimeOracle MAA Best PracticesPlannedDowntime
- 40. System Failures: Are the result of hardware failures, power failures, and operating system or server crashes.Data Failures: It is vital to design a solution to protect against, and recover from, data and media failure.
- 41. Database Replay
- 43. Partitioning
- 48. More Concurrency
- 49. Result CacheDisaster Recovery: Oracle Data Guard is a high availability and data protection capability optimized for the oracle database implemented from oracle 7 at 1992. Microsoft Database Mirroring has similar goals as Data Guard for increasing database availability and providing disaster protection, though there are many differences in its capabilities with Data Guard.Disaster Recovery(Cont)
- 51. Human Errors: Data failures can result from human error , which may be due to accidents or event sabotage.System Maintenance: As business needs change, system changes may also be required.Data Maintenance: As business requirements and processes change, the underlying data has to be maintained and transformed to suit the new environment, and done in such a way that there are minimal or no disruptions to the buesiness.Automatic Storage Management (ASM): ASM is a volume manager and a file system for Oracle database files that supports single-instance Oracle Database and Oracle Real Application Clusters (Oracle RAC) configurations. ASM is Oracle's recommended storage management solution that provides an alternative to conventional volume managers, file systems, and raw devices.
- 52. Features & CapabilitiesASM DiskgroupMirroringStripingStripingFailgroup1Failgroup2
- 53. Files, Extents, and Failure GroupsFiles and extent pointersFail groups and ASM mirroring
- 55. ASM and Oracle Managed Files
- 56. ASM Database
- 57. ASM Components
- 59. Performance
- 60. UtilitiesASM and Oracle Managed Files
- 61. ASM extends the power of Oracle Managed Files
- 62. Creates and manages Files automatically
- 63. Understands different Oracle file types
- 64. Offers built-in data mirroring and striping Capabilities
- 65. Automatically balances I/O load across disk group(s)
- 66. Functionally like as LVMASM Database
- 67. When ASM is enabled, Oracle automatically creates a small database for storing ASM-related meta-data
- 68. The default database name is “+ASM”
- 69. The default RAC instances are “+ASMn”, where n is the node number
- 70. The ASM database may be managed with srvctl commands
- 71. Normally auto-managed by OracleASM Components
- 72. Disk Groups
- 73. Oracle may provide disk mirroring (optional)
- 74. Oracle will stripe files across all available disks
- 75. Using ASM striping across a group of hardware striped LUNs can be beneficial to performance
- 76. Disks assigned to ASM Disk groups become ASM Disks
- 77. Files written on ASM disks become ASM files
- 78. Files are named with an Oracle-defined format
- 79. File name aliases may be defined for convenience
- 80. A hierarchical directory structure may be defined for file aliasesASM Components(Cont…)Templates may be used to associate file attributes such as redundancy and striping with ASM files
- 81. ASM is implemented through an ASM instance that runs as a service on each nodeDisk Group Redundancy
- 82. ASM Disk Groups may be defined with different levels of redundancy
- 83. NORMAL redundancy indicates that Oracle will create a software mirror for each disk
- 84. HIGH redundancy indicates that Oracle will create two mirror disks per data disk
- 85. Disk mirrors may be segregated into different locations by designating Failure Groups
- 86. EXTERNAL redundancy indicates that Oracle will not create any software mirrors. Hardware mirrors (if any) will be used instead
- 87. External redundancy is recommended for any system with a highly available storage arrayPerformance
- 88. ASM performance is sensitive to the number of disk spindles utilized
- 89. Just like traditional storage, the number of IOs/second that can be sustained is directly proportional to the number of disks in the disk group
- 90. Always use External Redundancy if you are using RAID-capable storage hardware
- 91. Multiple RAID 1 disks can be utilized in a Disk Group
- 92. ASM will provide the disk striping
- 93. Another option is to use multiple four disk RAID 10 groups as the building block for ASM Disk Groups
- 94. ASM will stripe on top of the hardware stripe, creating a stripe “plaid”, very desirable for OLTP and random I/O performanceUtilities
- 95. ACMCMD
- 96. SQLPLUS
- 97. OEMOracle DataGuard: An application-transparent high-performance low-impact asymmetrical online reliable Redo or SQL level background standby database transaction exchange utility capable of reporting, switchover and Failover.Data Guard helps you protect your Data.
- 98. Takes your data and automatically puts it elsewhere
- 99. Makes it available for Failover in case of failure.
- 100. The other capabilities are pure bonus.
- 102. Reporting
- 103. Off-loading Queries
- 104. BackupsData Guard PyramidEM Data Guard Manager Broker and CLI D a t a G u a r dProduction / Primary DatabasesPhysical and Logical Standby Databases
- 105. High Level
- 106. Data Guard comprises of two parts
- 107. REDO APPLY (DR)
- 108. Maintains a physical, block for block copy of the Production (also called Primary) database.
- 110. Maintains a logical, transaction for transaction copy of the Production database.Data Guard Redo Apply: Best for DRData Guard BrokerPhysical StandbyDatabasePrimaryDatabaseOptionalDelayBackupNetworkRedo ApplySync or Async Redo ShippingPhysical Standby Database is a block-for-block copy of the primary database
- 111. Uses the database recovery functionality to apply changes
- 112. Can be opened in read-only mode for reporting/queries
- 113. Can also perform backup, offloading production database
- 114. The best solution for DRData Guard SQL ApplyAdditionalIndexes & Materialized ViewsData Guard BrokerPrimaryDatabaseLogical StandbyDatabaseOptional DelayContinuouslyOpen for ReportsNetworkSync or Async Redo ShippingTransform Redo to SQL and ApplyLogical Standby Database is an open, independent, active database
- 115. Contains the same logical information (rows) as the production database
- 116. Physical organization and structure can be very different
- 117. Can host multiple schemas
- 118. Can be queried for reports while logs are being applied via SQL
- 119. Can create additional indexes and materialized views for better query performance
- 120. Not all Data Types supported (See the manual for a list)ReportingBackupsTapeStandbyDatabaseStandby Databases Are Not IdleStandby ServerStandby database can be used to offload the primary database, increasing the ROI
- 121. Standby SitePrimary SiteStandbyDatabaseProductionDatabaseOptional Delayed ApplyProtection from Human Errors and Data Corruptions
- 122. The application of changes received from the primary can be delayed at standby to allow for the detection of user errors and prevent standby to be affected
- 123. The apply process also revalidates the log records to prevent application of any log corruptionsOracle 11g Top Features:
- 124. Database Replay
- 126. Partitioning
- 129. SQL Plan Management
- 130. Private Statistics
- 131. More Concurrency
- 132. Result CacheDatabase ReplayCapture directoryShadow capture fileShadowShadowShadowShadowShadow capture fileRecording infrastructureShadow capture fileDatabase stackShadow capture fileBackgroundBackgroundDatabasebackupProductiondatabase
- 133. Database Replay(Cont) Oracle Database Replay addresses the issues associated with environmental changes by providing the ability to test the impact of those changes on a test system. Thus, you can gauge the impacts of these changes before you move them to production. The testing takes a sample workload that you record during a given period of time and “replays” it on your test system. Important workload attributes such as concurrency and transactional dependencies are maintained to make the testing as real-world as possible. What kinds of changes might you test? Such changes include:Database Upgrade
- 135. RAC related changes(adding nodes, interconnect changes, …)
- 136. OS platform changes and upgrades
- 137. Hardware changesSQL Performance Analyzer:It’s very similar to Database Replay; but with some important differences. SPA doesn’t record anything; you can capture the SQL statements from the SQL Tuning Sets or the library cache, or you can write your own SQL statements. Once it gets a number of SQL statements, it replays them against the database after changing some parameter you specify.ClientClientClient DBTestProductionStorage DBStorageCapture SQLExecute SQLMake ChangeExecute SQLCompare Perf
- 138. SQL Performance Analyzer EM Support:
- 139. Partitioning:Several New partitioning features introduced in Oracle Database 11g provide IT organization with simplicity in managing their partitioned objects while increasing their ability to scale and perform. Below is a matrix of all partition types.Transparent Tablespace Encryption: Oracle Database 10g introduced Transparent Data Encryption (TDE), allowing you to encrypt data within specific columns of a table. Oracle Database 11g enhances Transparent Data Encryption by providing for the encryption of all contents of a given Tablespace. Any permanent Tablespace can be encrypted. Undo and temporary tablespaces cannot be encrypted; however, all blocks and data created in these Tablespace types will be encrypted if they originate from an encrypted Tablespace. Additionally the redo log stream will be encrypted.
- 141. 3DES168
- 142. AES128
- 143. AES192
- 144. AES256Transparent Tablespace Encryption: Oracle Database 10g introduced Transparent Data Encryption (TDE), allowing you to encrypt data within specific columns of a table. Oracle Database 11g enhances Transparent Data Encryption by providing for the encryption of all contents of a given Tablespace. Any permanent Tablespace can be encrypted. Undo and temporary tablespaces cannot be encrypted; however, all blocks and data created in these Tablespace types will be encrypted if they originate from an encrypted Tablespace. Additionally the redo log stream will be encrypted.
- 146. 3DES168
- 147. AES128
- 148. AES192
- 149. AES256Flashback Data Archive:In Oracle 9i flashback transactions has been introduced. When data is updated the past image of the block is stored in undo segments, even if the data is committed. But mentioned data came from undo segments and undo data is not kept forever it is flushed if it’s not needed and there is a demand for space from a new transactions. Oracle 11g a new feature called Flashback Data Archive introduced. Select * from orders AS OF ‘Midnight 31-Dec-2004’ORDERSArchiveTablesChangesUser TablespacesFlashback Data ArchiveOracle Database
- 150. Flashback Data Archive(Cont)Define a flash back and add a table to that:
- 151. Create flashback archive far1 tablespace ts_far1 quota 1g retention 1 year;
- 152. Alter table test flashback archive far1;
- 153. By adding table test to flashback archive, whenever the table is changed, the past image will be stored in the flash recovery area.
- 154. Flashback Data Archive advantage over triggers:
- 155. The process done by oracle software code and oracle internal process it self. So there is no trigger related performance impact , no context-switching or checking for dependencies.
- 156. It’s possible to continue using the queries for flashback queries and limitation related to undo segments has been rectified.SQL Plan Management:What is your biggest concern while collecting statistics on objects? Is it the fear of breaking something , and more specifically, breaking a perfect execution plan? In oracle 11g this problem resolved through a feature called SQL Plan Management (SPM). Whenever a new execution plan is calculated , it is base lined in a store called SQL Management Base. Once base lined, the SQL execution plan is attached to the query.
- 157. To enable this feature we should set below parameter to TRUE:
- 158. Optimizer_capture_sql_paln_baselinesPrivate Statistics:Oracle 11g allows DBMS_STATS to gather stats in private area instead of storing in Data Dictionary. This way oracle Optimizer continue to work with stats collected in Data Dictionary until you are done in validating the new Stats collected in Private area. When Validating of statistics completed it can be published in Data Dictionary.More ConcurrencyAdd more columns or modify a column it can fail by below error: ORA-00054: resource busy and acquire with NOWAIT specified Above error occur because table has been locked by some transactions, so the other sessions can’t get an exclusive lock on the table to execute the DDL statement.
- 159. In Oracle Database 11g there are two major enhancement to improve concurrencies like these:
- 160. You can add a column to the table event when another transaction holds row locks on the table
- 161. You can instruct the session to executing the DDL to wait for some specified time in seconds.
- 162. Alter session set ddl_lock_timeout=30;123Result CacheWith SQL Query Result Cache the results of a query are stored in the SGA , and then can be used in subsequent queries.
- 163. To use of this feature:
- 164. Set the result_cache_max_size SQLQuery Result CacheSELECT …SELECT …Session 1Session 2
- 165. SQL Server Migration To Oracle: An Oracle database provides you with better scalability, reliability, increasedperformance, and better security than other third-party databases. For this reason,organizations migrate from their current database, such as Microsoft SQL Serverand Microsoft Access, to an Oracle database.