- 1、本文档共14页,可阅读全部内容。
- 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
oracle设计流程(国外英语资料)
oracle设计流程(国外英语资料)
Design of super large ORACLE database application system
Editor: admin Update Date: 2005-8-6
I. Introduction
The characteristics of ultra large systems are:
1., the number of users generally more than one million, and some more than 10 million, the amount of data in the database generally more than 1TB;
2., the system must provide real-time response function, the system does not stop running, requiring high availability and scalability of the system
In order to achieve the above requirements, in addition to the need for superior performance computers and mass storage equipment, but also the need for advanced data
The general ultra large system adopts dual or multi machine cluster system. The following database uses ORACLE 8.0.6 parallel server
Determine the systems ORACLE parallel server application partitioning strategy
Physical structure design of database
System hard disk partitioning and allocation
Backup and recovery policy considerations
Two, ORACLE parallel server application partition strategy
The ORACLE parallel server allows multiple INSTANCE instances on different nodes to access a database at the same time to improve the system
Availability, scalability, and performance of. Each INSTANCE instance in the ORACLE parallel server can be shared in the shared database
The table or index block is read into the local buffer, which means that a data block can exist in multiple INSTANCE instances
SGA district. Then it is important to keep the consistency of the data in these buffers. ORACLE uses PCM (Parallel)
Cache Management) locks maintain the consistency of buffers, while ORACLE manages through I DLM (integrated distributed locks)
PCM lock, and through the specialized LCK process to achieve data between the INSTANCE instance consistent.
Consider this situation: INSTANCE1 modifies the BLOCK X block, and then the INSTANCE2 needs to be modified for the BLOCK X block.
The ORACLE parallel server makes use of the PCM lock mechanism to write BLOCK
文档评论(0)