Cloud Computing in Government Environments USPTO Perspective Mr. Toby Bennett Acting Director, Office of Policy & Governance Office of the Chief Information Officer United States Patent and Trademark Office
Agenda Background US Federal Cloud Computing Strategy Cloud Strategy Considerations USPTO Perspective Current Approach Questions 9/10/2012 2
Cloud Computing Background Service - provides compute resources over a network Utility model sharing of resources across multiple users / pay for what is used Evolved from several underlying technologies Commodity computing, network, & storage; virtualization; service oriented architecture Private data centers highly underutilized (10%) 2006: First commercial providers of cloud services In 2011, US Government issues Cloud First Policy Recognizes cloud computing - fundamental technology shift 9/10/2012 3
Federal Cloud Computing Strategy Our responsibility in government is to achieve the significant cost, agility and innovation benefits of cloud computing as quickly as possible Each agency will evaluate its technology sourcing strategy so that cloud computing options are fully considered, consistent with the Cloud First policy. Vivek Kundra, U.S. Chief Information Officer, White House 9/10/2012 4
Cloud Strategy Framework Decision Framework for Cloud Computing 9/10/2012 5
Cloud Strategy Considerations For Federal Agencies What are the advantages / tradeoffs for various service models (SaaS, PaaS, IaaS) and deployment models (public, community, private, hybrid) based on: Size, utilization, cost of existing infrastructure investments including associated management overhead Scalability / elasticity needs (gradual or on-demand provisioning) Type of software investments (COTS versus in-house) & development approach Characteristics of IT operations (central or distributed user base, network bandwidth & reliability) Data integration / synchronization strategy IT security requirements (data sensitivity, federal regulations, auditing / continuous monitoring) Availability requirements / Continuity of operations plans Application & data portability 9/10/2012 6
USPTO Perspective Business Operations Over $2.4 billion in Annual Revenue Main complex located in Alexandria, VA Satellite Office in Detroit new openings planned for Dallas, Denver, & Silicon Valley 11,000+ Employees 3 Unions / Collective Bargaining Agreements Large population of teleworkers 82+ Patent & Trademark Resource Centers throughout the United States High volume of production Received 500,000+ patent applications Granted 240,000+ patents Received 395,000+ trademark applications Registered 175,000+ trademarks Reassigned 450,000+ patent and trademark rights Delivered 550,000+ patent and trademark information products Significant data sharing with international IPO partners 9/10/2012 7
USPTO Perspective IT Operations $400+ million in IT expenditures in Fiscal Year 2013 ~ 60% operations / maintenance ~ 40% improvement projects Investment to O&M ratio has been improving since 2010 Main data center located in Alexandria, VA 1561 Physical servers Support 3200 virtual server instances 3.5 Petabytes storage Shortening time windows for performing system maintenance Currently Weekdays 5.5 hrs. Weekends 11 hrs. Backup data center in PA IT Employees: 460+ government; 870+ contractors Significant amount of legacy software ~ 200 distinct applications / automated information systems Many are specific to Patents / Trademarks examination Developed by USPTO over decades using various technologies 9/10/2012 8
USPTO Perspective USPTO Business Considerations are Very Complex! 9/10/2012 9
USPTO Evolutionary Approach to Cloud Computing 2008: USPTO started 5 year effort to revitalize IT infrastructure Standardize / virtualize data center computing Upgrade network bandwidth 10X (WAN & LAN) Improve development methodologies 2010: USPTO began efforts studying cloud computing Started development on first public cloud application as pilot (TDRS 2.0) 2011: USPTO launched two new major system investments Patents End-to-End (PE2E): Electronic end-to-end text based processing of patent applications Trademarks Next Generation (TMNG): virtualize legacy / build new value stream applications based on new business architecture Each based on cloud compliant technology stacks To support these investments, USPTO chose to implement a hybrid cloud approach. Transform existing data center to internal private cloud using open source technologies. Use API for external hybrid cloud provisioning. 9/10/2012 10
E a r l y A d o p t i o n USPTO Cloud Approach Data Center Evolution USPTO IT Infrastructure Perspective Our History - Evolution from Virtualization to Cloud Computing V M W a r e 2 0 0 6 : I n i t i a l v i r t u a l i z a t i o n o f L i n u x s e r v e r s l e v e r a g i n g e x i s t i n g V M w a r e e n v i r o n m e n t u s e d f o r V i r t u a l D e s k t o p I n f r a s t r u c t u r e ( V D I ) X E N 2 0 0 7 : A d o p t e d X e n h y p e r v i s o r a s a n o p e n s o u r c e a l t e r n a t i v e t o V M w a r e f o r L i n u x s e r v e r v i r t u a l i z a t i o n K V M 2 0 0 9 : A d o p t e d K V M, a n o p e n s o u r c e h y p e r v i s o r w i t h s u p e r i o r p e r f o r m a n c e, m a n a g e a b i l i t y, a n d s e c u r i t y C u r r e n t S t a t e R H E V 2 0 1 0 : A d o p t e d R H E V, a v i r t u a l i z a t i o n m a n a g e m e n t s o l u t i o n f o r o u r e x i s t i n g K V M e n v i r o n m e n t C l o u d F o r m s 2 0 1 2 : A d o p t e d C l o u d F o r m s, a h y b r i d c l o u d m a n a g e m e n t s o l u t i o n c o m p a t i b l e w i t h o u r e x i s t i n g R H E V e n v i r o n m e n t 11
USPTO Cloud Approach Benefit Hybrid cloud open source approach: Private cloud Leverages existing data center investments Facilitates new software system investments Can develop & deliver new services incrementally (Agile methodology) Simplifies data integration / migration strategy Open source helps maintain application / data portability & avoids vendor lock-in Meets Federal IT security requirements Public cloud USPTO Perspective Deploy to public cloud in future to support growth, greater efficiency, higher peak capacity, higher availability (dynamic failover) 9/10/2012 12
USPTO Cloud Approach PE2E Plans & Progress Patent End to End (PE2E) Cloud Compliant Functionality Patent Application Text Initiative Convert patent application data from image to structured text [XML / ST.96]) Currently 63,000,000 pages of claims, specification, and abstracts are available to the entire Patent Corps. Provides analytical tools for claim tree analysis, content comparison, claim term analysis / phrase usage. Common Patent Classification systems (move to IPC compliant classification system shared with EPO) Patent Family Maps PE2E Examiner Interface Text2PTO Applicant Interface (online text based filing) Enables office to accept patent applications in text format rather than PDF. Allows new analytical reporting tools, automated formality checking, access to patent families/continuity/claim dependencies/other application content. 9/10/2012 13
USPTO Cloud Approach PE2E Platform Technologies Open Source Core Platform (Middleware): MySQL open source relational database JBoss - Java EE-based application server Solr Lucene based search server Subversion version control system Maven project development management & comprehension tool Ant - Java-based build tool Hudson - continuous integration tool 9/10/2012 14
USPTO Cloud Approach Patent Data Dissemination Information Dissemination The USPTO currently has an agreement with Google for public hosting and dissemination of USPTO bulk, electronic patent and trademark data at no charge to the public. Under the agreement USPTO provides the company with existing bulk, electronic files and access to daily, weekly, and bimonthly updates. Electronic files provided include images and text, totaling several Terabytes, of patent grants and published applications, trademark applications, patent classification information, patent and trademark assignments and associated documentation. The agreement allows USPTO to focus its resources on creation of the data files and rely upon the external public cloud to disseminate the information to thousands of customers. 9/10/2012 15
Questions 9/10/2012 16
政府环境中的云计算 美国专利 商标局透视 托比 班尼特先生美国专利商标局首席信息官办公室政策与治理办公室代理主任
议程 背景介绍 美国联邦云计算战略 云战略考虑 美国专利商标局透视 现行办法 提问 9/11/2012 2
云计算背景介绍 业务 通过网络提供计算资源 实用新型 多用户共享资源 / 付费使用 由几项底层技术进化而来 商品计算 网络 仓储 虚拟化 面向服务的架构 专用数据中心高度地未充分利用 (10%) 2006 年 : 第一批商用云服务供应商 2011 年, 美国政府颁布 云优先策略 确认云计算为根本技术变革 联邦战略指导各机构在新的 IT 投资方面重点考虑云 9/11/2012 3
美国联邦云计算战略 我们在政府的职责就是尽快获得云计算重大的成本 便捷及创新效益 每个机构都要评估其技术导入战略, 充分考虑云计算优先原则 遵循云优先策略 维维克. 昆德拉, 美国白宫联邦政府首席信息技术官 9/11/2012 4
云战略框架 云计算决策框架 9/11/2012 5
联邦政府各分支机构对于云计算 需要考虑的方面 各种服务模式 ( 软件即服务 平台即服务 基础设施即服务 ) 和部署模式 ( 公共 社区 私用 混合 ) 的优势或权衡的基础是什么 : 现有基础设施投资的规模 利用率 成本, 包括相关的管理开销 可扩展性和灵活性要求 ( 逐渐的或按需的供应 ) 软件投资类型 ( 商用货架产品与内部产品 ) 及开发途径 IT 运营特点 ( 中央式或分布式用户基础 网络带宽及可靠性 ) 数据整合及同步策略 IT 安全性要求 ( 数据灵敏性 联邦政府规范 审计与持续监控 ) 可用性需求及营运计划的连续性 应用及数据迁移 9/11/2012 6
美国专利商标局透视商业运营 年营业收入超过 24 亿美元 位于弗吉尼亚州亚历山大市的主要综合设施 底特律的卫星办公室 为达拉斯 丹佛及硅谷规划的新站 11,000 多员工 3 个工会或劳资谈判合同 庞大的远程工作人员数 全美国超过 82 个专利和商标资源中心 高产量 收到 500,000 多份专利申请书 批准 240,000 多项专利 收到 395,000 多份商标申请 注册了 175,000 多份商标 重新分配了 450,000 多项专利和商标权 交付了 550,000 多件专利和商标权信息产品 与国际知识产权机构伙伴的重要数据共享 9/11/2012 7
美国专利商标局透视 IT 运营 2013 年财政年度 IT 支出 4 亿多美元 ~ 60% 用于运维 ~ 40% 用于改进项目 2010 年以来投资运维的比率一直在提高 位于弗吉尼亚州亚历山大市的主数据中心 1561 个物理服务器 支持 3200 个虚拟服务器实例 3.5 千兆字节存储 缩短系统维护时间窗口 目前是工作日 5.5 小时, 周末 11 小时 宾夕法尼亚州的备份数据中心 IT 员工 :460 多在政府部门 ;870 多个承包商 大量的遗产软件 ~ 200 个不同的应用程序和自动化信息系统 许多是针对专利或商标审查的 由美国专利商标局利用各种技术历经数十年开发的 9/11/2012 8
美国专利商标局透视 美国专利商标局商业方面的考虑极其复杂!! 9/11/2012 9
美国专利商标局云计算的发展途径 2008 年 : 美国专利商标局开始了 5 年的努力以恢复 IT 基础设施 标准化 / 虚拟化数据中心计算 升级网络带宽 10X ( 广域网 & 局域网 ) 改进开发方法 2010 年美国专利商标局开始努力研究云计算 开始在第一个公共云应用上进行试点开发 ( 跟踪与数据中继卫星 2.0) (TDRS 2.0) 2011: 年美国专利商标局启动两项新的主系统投资 专利端到端 (PE2E): 基于电子端到端文本的专利申请处理 商标新一代 (TMNG): 虚拟化遗产 / 基于新的业务架构打造新的价值流应用 每一项都基于云兼容的技术堆栈 为支持这两项投资, 美国专利商标局选择实施混合式云途径 使用开放源代码技术将现有数据中心转变为内部专用云 采用应用程序接口向外部提供混合云 9/11/2012 10
E a r l y A d o p t i o n 美国专利商标局云途径 数据中心的发展 美国专利商标局 IT 基础设施透视 我们的历史 从虚拟化向云计算的演进 V M W a r e 2 0 0 6 : I n i t i a l v i r t u a l i z a t i o n o f L i n u x s e r v e r s l e v e r a g i n g e x i s t i n g V M w a r e e n v i r o n m e n t u s e d f o r V i r t u a l D e s k t o p I n f r a s t r u c t u r e ( V D I ) X E N 2 0 0 7 : A d o p t e d X e n h y p e r v i s o r a s a n o p e n s o u r c e a l t e r n a t i v e t o V M w a r e f o r L i n u x s e r v e r v i r t u a l i z a t i o n K V M 2 0 0 9 : A d o p t e d K V M, a n o p e n s o u r c e h y p e r v i s o r w i t h s u p e r i o r p e r f o r m a n c e, m a n a g e a b i l i t y, a n d s e c u r i t y C u r r e n t S t a t e R H E V 2 0 1 0 : A d o p t e d R H E V, a v i r t u a l i z a t i o n m a n a g e m e n t s o l u t i o n f o r o u r e x i s t i n g K V M e n v i r o n m e n t C l o u d F o r m s 2 0 1 2 : A d o p t e d C l o u d F o r m s, a h y b r i d c l o u d m a n a g e m e n t s o l u t i o n c o m p a t i b l e w i t h o u r e x i s t i n g R H E V e n v i r o n m e n t 11
混合云开放源代码途径 : 专有云 调节现存数据中心投资 促进新软件系统投资 可以逐渐开发和交付新业务 ( 灵活的方法 ) 简化数据整合 / 迁移策略 开放源代码有助于维护应用 / 数据迁移和避免厂商锁定 符合联邦政府 IT 安全要求 公共云 美国专利商标局云途径效益 美国专利商标局透视 给今后的公共云部署更高效 更高峰值容量 更高可用性 ( 动态故障切换 ) 以支持增长 电子邮件和协作服务考虑公共软件即服务模式 9/11/2012 12
美国专利商标局 云途径专利端到端规划与进展 专利申请书主动生成功能 将专利申请数据从镜像文件转换到结构化文本 ([XML / ST.96]) 目前整个专利公司可提供 63,000,000 页专利权要求 说明和摘要 提供专利权要求树分析 内容对比 权利要求术语分析 / 短语用法分析等分析工具 常见的专利分类系统 ( 可到达与欧洲专利组织共享的国际专利分类的兼容分类系统 ) 专利族地图 专利端到端 (PE2E) 云兼容功能 专利端到端 (PE2E) 检验接口 文本到专利局申请接口 ( 基于在线文本的申请 ) 使专利局可以接受文本格式而非 PDF 格式的专利申请书. 支持新的分析性报表工具 自动化手续检验 访问专利族 / 连续性 / 权利要求相关性 / 及其他申请内容 9/11/2012 13
美国专利商标局 云途径专利端到端平台技术 开放源代码核心平台 ( 中间件 ): MySQL 开放源代码的关系数据库 JBoss - Java 基于 EE 的应用服务器 Solr 基于 Lucene 的搜索应用服务器 Subversion 版本控制系统 Maven 项目开发管理及理解工具 Ant 基于 Java 的构建工具 Hudson - 持续集成工具 9/11/2012 14
美国专利商标局 Cloud Approach Patent Data Dissemination Information Dissemination 美国专利商标局目前与谷歌达成协议向公众免费托管和发布美国专利商标局大量的电子专利和商标数据 根据该协议, 美国专利商标局向谷歌公司提供现有的大量电子文件和访问每日 每周 双月更新的权限 所提供的电子文件包括专利许可 公开的专利申请 商标申请 专利分类信息 专利和商标分配及相关文件的图像和文本, 总计达几兆兆字节 该协议使美国专利商标局可以集中其资源创建数据文件并通过外部公共云向成千上万的客户发布信息 9/11/2012 15
提问 9/11/2012 16