Data Lifecycle Review and Outlook - Luca Canali - CERN数据生命周期的回顾与展望——卢卡康纳利-欧洲核子研究中心.pptVIP

下载本文档

4
0
约6.44千字
约 20页
2018-04-16 发布于江西
举报
版权申诉

Data Lifecycle Review and Outlook - Luca Canali - CERN数据生命周期的回顾与展望——卢卡康纳利-欧洲核子研究中心.ppt

1、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。。
2、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。
3、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)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

Data Lifecycle Review and Outlook - Luca Canali - CERN数据生命周期的回顾与展望——卢卡康纳利-欧洲核子研究中心.ppt

DB Communication- * CERN IT Department CH-1211 Genève 23 Switzerland www.cern.ch/it CERN IT Department CH-1211 Genève 23 Switzerland www.cern.ch/it CERN IT Department CH-1211 Genève 23 Switzerland www.cern.ch/it DB Communications - * Data Lifecycle Review and Outlook Luca Canali, CERN Distributed Database Operations Workshop April 20, 21 Barcelona Outline Motivations and main problems addressed by datalife cycle management Implementation ideas and techniques for the physics DBs Sharing results from production Ideas for future work Data Lifecycle Management Motivated by large data volumes produced by LHC experiments Large amounts of data will be collected and stored for several years Discussion started in November 2008 Main issues of databases that grow ‘too large’ Administration Performance Cost Administration of VLDB Administration and performance of VLDBs can suffer from Full table/partition/index scan Data reorganization and/or bulk delete DB-wide operations (restore, stats gathering) VLDB advantages: Data consolidation, application consolidation Task Identify how to get advantages of consolidation coexist and a ‘lean’ DBs for manageability and performance Active Datasets Many applications are structure as write-once read-many At a given time typically only a subset of data is actively used Natural optimization: having large amounts of data that are set read only Can be used to simplify administration Replication and backup can profit too Problem Not all app are ready for this type of optimization Time-Organized data Several key database tables are naturally time organized this leads to range-based partitioning Other solution is ‘manual split’ i.e. multiple similar tables in different schemas Advantages Partitions can be treated as separate tables for bulk operations Full scan operation, if they happen, do not span all tables Partitioning Issues Index strategy Indexes need to be local partitioned in the ideal case to fully make use of ‘partition i