
平台论坛博客文库

论坛徽章:: 7

41楼 [报告]

发表于 2012-07-05 08:25 |只看该作者

59355189 发表于 2012-05-11 15:51
销售也可以帮忙的嘛，技术团队出的方法也比较好符合要求

销售的点子离实际实现其实还是挺远的，销售可以把一个烂产品吹的天花乱坠。
根据自己的容灾经验，个人觉得比较现实的两地三中心的方案还是本地两点做RAC，保证本地两点的数据完全一致。异地基于数据库日志作复制，使用SHAREPLEX或者GOLDENGATE，允许小部分数据丢失。

实战分享：从技术角度谈机器学习入门| 【大话IT】RadonDB低门槛向MySQL集群下战书 | ChinaUnix打赏功能已上线！ | 新一代分布式关系型数据库RadonDB知多少？

spook

富足长乐

论坛徽章:: 1

42楼 [报告]

发表于 2012-07-05 08:34 |只看该作者

milujite 发表于 2012-07-05 08:25
销售的点子离实际实现其实还是挺远的，销售可以把一个烂产品吹的天花乱坠。
根据自己的容灾经验，个人 ...

你是在说我吗？

你在两个距离10多公里的数据中心各放一个机器做RAC的方案很高明啊，网络啊，链路啊，带宽啊什么都不考虑，你真是做技术的吗？

mike79

大富大贵

论坛徽章:: 12

43楼 [报告]

发表于 2012-07-05 09:10 |只看该作者

milujite 发表于 2012-07-05 08:25
销售的点子离实际实现其实还是挺远的，销售可以把一个烂产品吹的天花乱坠。
根据自己的容灾经验，个人 ...

extend rac这个方案你实施过么？在实施和维护中要注意哪些问题？异地第三站点有哪些要求？

wpscn

白手起家

论坛徽章:: 0

44楼 [报告]

发表于 2012-07-05 14:09 |只看该作者

milujite 发表于 2012-07-05 08:25
销售的点子离实际实现其实还是挺远的，销售可以把一个烂产品吹的天花乱坠。
根据自己的容灾经验，个人 ...

你牛x，国内有人这么干过，结局很惨。最终失败告终。又是一个销售出身

milujite

小富即安

论坛徽章:: 7

45楼 [报告]

发表于 2012-07-05 15:34 |只看该作者

本帖最后由 milujite 于 2012-07-05 15:44 编辑

spook 发表于 2012-07-05 08:34
你是在说我吗？

你在两个距离10多公里的数据中心各放一个机器做RAC的方案很高明啊，网络啊，链路啊， ...

sfrac我是没做过，但是goldengate做过很多，虽然有点坑爹，但是技术上是可行的，对数据库设计者要求很高。能做的起两地三中心的，如果链路都不能搞定，用不起，还做毛容灾。直接卡车运算了

milujite

小富即安

论坛徽章:: 7

46楼 [报告]

发表于 2012-07-05 15:48 |只看该作者

本帖最后由 milujite 于 2012-07-05 15:48 编辑

wpscn 发表于 2012-07-05 14:09
你牛x，国内有人这么干过，结局很惨。最终失败告终。又是一个销售出身

能讲讲失败在哪里么？我不是销售，容灾方面经验有限，做过存储复制和GOLDENGATE等。

spook

富足长乐

论坛徽章:: 1

47楼 [报告]

发表于 2012-07-06 11:00 |只看该作者

milujite 发表于 2012-07-05 15:34
sfrac我是没做过，但是goldengate做过很多，虽然有点坑爹，但是技术上是可行的，对数据库设计者要求很高 ...

【根据自己的容灾经验，个人觉得比较现实的两地三中心的方案还是本地两点做RAC，保证本地两点的数据完全一致。异地基于数据库日志作复制】

哥们儿，你确认两地三中心中间的本地双中心和 RAC 的中的本地（local）是一个概念吗？你用 goldengate 做的本地两点数据完全一致？你确认你用的不是 dataguard ？你想想为什么楼下会想到 extend RAC ……

milujite

小富即安

论坛徽章:: 7

48楼 [报告]

发表于 2012-07-06 11:20 |只看该作者

本帖最后由 milujite 于 2012-07-06 11:33 编辑

spook 发表于 2012-07-06 11:00
【根据自己的容灾经验，个人觉得比较现实的两地三中心的方案还是本地两点做RAC，保证本地两点的数据完全 ...

我心中的方案是类似SFRAC做同城，配合GOLDENGATE做异地。你哪里看到我说goldengate能够保证数据完全一致了，有点常识的人都知道异地长距离一致不现实，goldengate的优点就是备节点可用，并且备节点能够用于查询等业务，goldengate的单位是记录，而不是数据块。我说的数据一致是指同城OK，缺点也是肯定的，数据库要调整/优化/规范。如果你认为这方案什么问题，大可说出这方案缺点在哪里，哪个方案都是肯定有缺点的，我很愿意去倾听学习。

wolfop

大富大贵

论坛徽章:: 5

49楼 [报告]

发表于 2012-07-06 13:11 |只看该作者

milujite 发表于 2012-07-06 11:20
我心中的方案是类似SFRAC做同城，配合GOLDENGATE做异地。你哪里看到我说goldengate能够保证数据完全一致了， ...

最糟糕的方案，放着MAA建议不用。

milujite

小富即安

论坛徽章:: 7

50楼 [报告]

发表于 2012-07-06 14:41 |只看该作者

本帖最后由 milujite 于 2012-07-06 15:02 编辑

wolfop 发表于 2012-07-06 13:11
最糟糕的方案，放着MAA建议不用。

楼主要的是“两地三中心"，MAA是一个概念，在ORACLE的数据库上的实现是下面这样？

Oracle® Database High Availability Best Practices
10g Release 2 (10.2)
Part Number B25159-01
Home
Book List
Index
Master Index
Contact Us
Next
PDF · Mobi · ePub
Contents
List of Figures
List of Tables
Title and Copyright Information
Preface
Audience
Documentation Accessibility
Related Documents
Conventions
1 Introduction to High-Availability Best Practices
1.1 Oracle Database High-Availability Architecture
1.2 Oracle Database High-Availability Best Practices
1.3 Oracle Maximum Availability Architecture
1.4 Operational Best Practices
2 Configuring for High-Availability
2.1 Configuring Storage
2.1.1 Evaluate Database Performance Requirements and Storage Performance Capabilities
2.1.2 Use Automatic Storage Management (ASM) to Manage Database Files
2.1.3 Use a Simple Disk and Disk Group Configuration
2.1.4 Use Disk Multipathing Software to Protect from Path Failure
2.1.5 Use Redundancy to Protect from Disk Failure
2.1.6 Consider HARD-Compliant Storage
2.2 Configuring Oracle Database 10g
2.2.1 Requirements for High Availability
2.2.1.1 Enable ARCHIVELOG Mode
2.2.1.2 Enable Block Checksums
2.2.2 Recommendations for High Availability and Fast Recoverability
2.2.2.1 Configure the Size of Redo Log Files and Groups Appropriately
2.2.2.2 Use a Flash Recovery Area
2.2.2.3 Enable Flashback Database
2.2.2.4 Use Fast-Start Fault Recovery to Control Instance Recovery Time
2.2.2.5 Enable Database Block Checking
2.2.2.6 Set DISK_ASYNCH_IO
2.2.2.7 Set LOG_BUFFER to At Least 8 MB
2.2.2.8 Use Automatic Shared Memory Management
2.2.2.9 Increase PARALLEL_EXECUTION_MESSAGE_SIZE
2.2.2.10 Tune PARALLEL_MIN_SERVERS
2.2.2.11 Disable Parallel Recovery
2.2.3 Recommendations to Improve Manageability
2.2.3.1 Use Automatic Performance Tuning Features
2.2.3.2 Use a Server Parameter File
2.2.3.3 Use Automatic Undo Management
2.2.3.4 Use Locally Managed Tablespaces
2.2.3.5 Use Automatic Segment Space Management
2.2.3.6 Use Temporary Tablespaces and Specify a Default Temporary Tablespace
2.2.3.7 Use Resumable Space Allocation
2.2.3.8 Use Database Resource Manager
2.3 Configuring Oracle Database 10g with RAC
2.3.1 Connect to Database using Services and Virtual Internet Protocol (VIP) Address
2.3.2 Use Oracle Clusterware to Manage the Cluster and Application Availability
2.3.3 Use Client-Side and Server-Side Load Balancing
2.3.4 Mirror Oracle Cluster Registry (OCR) and Configure Multiple Voting Disks
2.3.5 Regularly Back Up OCR to Tape or Offsite
2.3.6 Verify That CRS and RAC Use Same Interconnect Network
2.3.7 Configure All Databases for Maximum Instances in the Cluster
2.4 Configuring Oracle Database 10g with Data Guard
2.4.1 Physical or Logical Standby
2.4.1.1 Benefits of a Physical Standby Database
2.4.1.2 Benefits of a Logical Standby Database
2.4.1.3 Determining Which Standby Type Is Best for Your Application
2.4.2 Data Protection Mode
2.4.3 Number of Standby Databases
2.4.4 General Configuration Best Practices for Data Guard
2.4.4.1 Enable Flashback Database for Easy Reinstantiation After Failover
2.4.4.2 Use FORCE LOGGING Mode
2.4.4.3 Use Data Guard Broker
2.4.4.4 Use a Simple, Robust Archiving Strategy and Configuration
2.4.4.5 Use Standby Redo Logs and Configure Size Appropriately
2.4.4.6 Parameter Configuration Example
2.4.5 Redo Transport Services Best Practices
2.4.5.1 Conduct Performance Assessment with Proposed Network Configuration
2.4.5.2 Best Practices for Primary Database Throughput
2.4.5.3 Best Practices for Network Configuration and Highest Network Redo Rates
2.4.6 Log Apply Services Best Practices
2.4.6.1 Redo Apply Best Practices for Physical Standby Databases
2.4.6.2 SQL Apply Best Practices for Logical Standby Databases
2.4.7 Role Transition Best Practices
2.4.7.1 Role Transition During Switchover
2.4.7.2 Role Transition During Failover
2.4.8 Maintaining a Physical Standby Database as a Clone
2.4.9 Recommendations on Protecting Data Outside of the Database
2.4.10 Assessing Data Guard Performance
2.5 Configuring Backup and Recovery
2.5.1 Use Oracle Database Features and Products
2.5.1.1 Use Recovery Manager to Back Up Database Files
2.5.1.2 Use Oracle Secure Backup
2.5.1.3 Use Restore Points
2.5.2 Configuration and Administration
2.5.2.1 Understand When to Use Backups
2.5.2.2 Determine a Backup Frequency
2.5.2.3 Use an RMAN Recovery Catalog
2.5.2.4 Enable Block Change Tracking for Incremental Backups
2.5.2.5 Enable Autobackup for Control File and Server Parameter File
2.5.3 Backup to Disk
2.5.3.1 Determine Disk Backup Methods
2.5.3.2 Create Backups in NOCATALOG Mode and RESYNC CATALOG Afterwards
2.5.3.3 Create Database Backups on Disk in the Flash Recovery Area
2.5.3.4 In a Data Guard Environment, Back Up to Flash Recovery Area on All Sites
2.5.4 Backup to Tape
2.5.4.1 Create Tape Backups from the Flash Recovery Area
2.5.4.2 Maintain Offsite Backups
2.5.5 Backup and Recovery Maintenance
2.5.5.1 Regularly Check Database Files for Corruption
2.5.5.2 Periodically Test Recovery Procedures
2.5.5.3 Regularly Backup the Recovery Catalog Database
2.6 Configuring Fast Application Failover
2.6.1 Configuring Clients for Failover
2.6.2 Client Failover in a RAC Database
2.6.3 Failover from a RAC Primary Database to a Standby Database
3 Monitoring Using Oracle Grid Control
3.1 Overview of Monitoring and Detection for High Availability
3.2 Using Oracle Grid Control for System Monitoring
3.2.1 Set Up Default Notification Rules for Each System
3.2.2 Use Database Target Views to Monitor Health, Availability, and Performance
3.2.3 Use Event Notifications to React to Metric Changes
3.2.4 Use Events to Monitor Data Guard System Availability
3.3 Managing the High-Availability Environment with Oracle Grid Control
3.3.1 Check Oracle Grid Control Policy Violations
3.3.2 Use Oracle Grid Control to Manage Oracle Patches and Maintain System Baselines
3.3.3 Use Oracle Grid Control to Manage Data Guard Targets
4 Managing Outages
4.1 Outage Overview
4.1.1 Unscheduled Outages
4.1.1.1 Managing Unscheduled Outages on the Primary Site
4.1.1.2 Managing Unscheduled Outages on the Secondary Site
4.1.2 Scheduled Outages
4.1.2.1 Managing Scheduled Outages on the Primary Site
4.1.2.2 Managing Scheduled Outages on the Secondary Site
4.1.2.3 Preparing for Scheduled Outages on the Secondary Site
4.2 Recovering from Unscheduled Outages
4.2.1 Complete Site Failover
4.2.1.1 When to Use Complete Site Failover
4.2.1.2 Best Practices for Complete Site Failover
4.2.1.3 Repair Solution
4.2.2 Database Failover with a Standby Database
4.2.2.1 When To Perform a Data Guard Failover
4.2.2.2 Best Practices for Implementing Data Guard Failover
4.2.3 Database Switchover with a Standby Database
4.2.3.1 When to Perform a Data Guard Switchover
4.2.3.2 Best Practices for Implementing Data Guard Switchover
4.2.4 RAC Recovery for Unscheduled Outages
4.2.4.1 Automatic Instance Recovery for Failed Instances
4.2.4.2 Automatic Service Relocation
4.2.4.3 Oracle Cluster Registry
4.2.5 Application Failover
4.2.6 ASM Recovery After Disk and Storage Failures
4.2.6.1 ASM Instance Failure
4.2.6.2 ASM Disk Failure
4.2.6.3 Data Area Disk Group Failure
4.2.6.4 Flash Recovery Area Disk Group Failure
4.2.7 Recovering from Data Corruption (Data Failures)
4.2.7.1 Use Data Guard to Recover From Data Corruption and Data Failure
4.2.7.2 Use RMAN Block Media Recovery
4.2.7.3 Use RMAN Data File Media Recovery
4.2.7.4 Re-Create Objects Manually
4.2.8 Recovering from Human Error
4.2.8.1 Resolving Table Inconsistencies
4.2.8.2 Resolving Row and Transaction Inconsistencies
4.2.8.3 Resolving Database-Wide Inconsistencies
4.3 Restoring Fault Tolerance
4.3.1 Restoring Failed Nodes or Instances in a RAC Cluster
4.3.1.1 Recovering Service Availability
4.3.1.2 Considerations for Client Connections After Restoring a RAC Instance
4.3.2 Restoring a Standby Database After a Failover
4.3.2.1 Restoring a Standby Database After a Fast-Start Failover
4.3.2.2 Reinstating a Standby Database Using Enterprise Manager After a Failover
4.3.3 Restoring ASM Disk Groups after a Failure
4.3.4 Restoring Fault Tolerance After Planned Downtime on Secondary Site or Clusterwide Outage
4.3.5 Restoring Fault Tolerance After a Standby Database Data Failure
4.3.6 Restoring Fault Tolerance After the Production Database Was Opened Resetlogs
4.3.7 Restoring Fault Tolerance After Dual Failures
4.4 Eliminating or Reducing Downtime for Scheduled Outages
4.4.1 Storage Maintenance
4.4.1.1 Migrating to ASM Storage
4.4.1.2 Adding and Removing Storage
4.4.2 RAC Database Patches
4.4.2.1 Best Practices To Minimize Downtime
4.4.3 Database Upgrades
4.4.3.1 Database Upgrade Assistant
4.4.3.2 Data Guard SQL Apply (Logical Standby)
4.4.3.3 Oracle Streams
4.4.3.4 Transportable Tablespaces
4.4.4 Database Platform or Location Migration
4.4.4.1 Transportable Database
4.4.4.2 Oracle Streams
4.4.4.3 Oracle Data Pump
4.4.4.4 Transportable Tablespaces
4.4.4.5 Data Guard Redo Apply (Physical Standby Database)
4.4.5 Online Database and Application Upgrades
4.4.6 Database Object Reorganization
4.4.7 System Maintenance
5 Migrating to an MAA Environment
5.1 Overview of Migrating to MAA
5.2 Migrating to RAC from a Single Instance
5.3 Adding a Data Guard Configuration to a RAC Primary
A Database SPFILE and Oracle Net Configuration File Samples
A.1 SPFILE Samples
A.2 Oracle Net Configuration Files
A.2.1 SQLNET.ORA Example for All Hosts Using Dynamic Instance Registration
A.2.2 LISTENER.ORA Example for All Hosts Using Dynamic Instance Registration
A.2.3 TNSNAMES.ORA Example for All Hosts Using Dynamic Instance Registration

复制代码

1 2 3 456 7 8 / 8 页下一页

返回列表

[容灾] 关于容灾方案的选择，请高手指教。 [复制链接]