An Overview of Grid, Cloud and related Database Technologies Xiaoming Gao Outline Grid technologies Cloud technologies Database technologies related to clouds Grid technologies Grid system structure Distributed file systems Resource management systems Reference: Ewa Deelman, et. al. Workflows and e-Science: An overview of workflow system features and capabilities. Distributed file systems
Reference: Tran Doan Thanh, et. al. A Taxonomy and Survey on Distributed File Systems. Distributed file system - Lustre Scale to petabytes of storage and hundreds of gigabytes of I/O throughput Single metadata server Failover mechanisms for metadata and object storage servers Reference: Sun Microsystems, Inc. Lustre File System white paper. Distributed file system - GPFS IBM General Parallel File System
Fully distributed architecture for both I/O and metadata operations Distributed locking management using tokens for concurrent data and metadata access Logging, failover, and replication mechanisms to handle node and disk failures Reference: Frank Schmuck, et. al. GPFS: A Shared-Disk File System for Large Computing Clusters. Grid resource management systems Resource types: computing resources, network resource, storage resources, service resources Resource management system abstract architecture: : Application to RMS interfaces : RMS to native operating system or hardware environment : Internal RMS functions Reference: Klaus Krauter, et. al. A Taxonomy and Survey of Grid Resource Management Systems. Grid resource management systems Design issues:
- Machine organization - Resource model, resource information storage, discovery and dissemination, QoS - Scheduler organization, scheduling policies System Grid type Organization Condor Computational Flat Grid Globus Grid Toolkit
query based discovery, periodic push dissemination Centralized scheduler Reference: Klaus Krauter, et. al. A Taxonomy and Survey of Grid Resource Management Systems. Higher-level tools (like Nimrod/G) and services offer scheduling support Grid workflow systems Definition of workflow: The automation of the processes, which involves the orchestration of a set of Grid
services, agents and actors that must be combined together to solve a problem or to define a new service. Workflow derivation: Dataflow model Workflow Distributed parallel programming Categories in terms of complexity: - Linear sequence of tasks - Directed acyclic graph - Cyclic graph: composition in space or distributed static dataflow - Compact graph: workflow of workflows - Implicit graph: workflows expressed as a set of desired outcomes Reference: Geoffrey Fox, et. al. Workflow in Grid Systems.
Grid workflow systems Workflow life cycle: - Composition, representation and execution model description - Mapping workflow to resources - Execution - Provenance in all stages of life cycle An example - Triana: - Compact graphical composition, Cyclic graph representation - External broker based resource mapping - Job level and service level execution, passive failure report, adaptive workflows - Detailed provenance recording Reference: Ewa Deelman, et. al. Workflows and e-Science: An overview of workflow system features and capabilities. Cloud technologies Definition of Cloud:
- A large pool of easily accessible virtualized resources - Dynamically scalable to a variable load, allowing for optimum resource utilization - Provided in a pay-per-use model, with QoS specified with SLAs Cloud Stack: Software as a Service Platform as a Service Infrastructure as a Service Reference: Luis M. Vaquero, et. al. A Break in the Clouds: Towards a Cloud Definition. Cloud technologies Contribution from Google: Hadoop MapReduce, Amazon Elastic MapReduce, Azure MapReduce MapReduce
Hadoop HBase, Amazon SimpleDB, Azure Table BigTable Google File System Hadoop Distributed File System Google File System
Targets at large files and write-once-read-many access styles Built on commodity hardware : failure as norm Files divided into fixed-size chucks and duplicated Logging, check-pointing and replication for fast recovery Reference: Sanjay Ghemawat, et. al. The Google File System. Google MapReduce framework MapReduce programming model: - map (k1, v1) -> list(k2, v2) - reduce (k2, list(v2)) -> list(v2) Applications: - Wordcount - Inverted index
- All-pair sequence alignment - K-means clustering (iterative MapReduce with Twister) - Google MapReduce framework Task rescheduling and master checkpoints to handle failures Backup tasks to deal with stragglers Reference: Jeffrey Dean, et. al. MapReduce: Simplified Data Processing on Large Clusters. BigTable Data Model: - A sparse, distributed, persistent multidimensional sorted map - A table can have multiple column families - A column family can have unbounded number of columns System design: -
Targeted at peta-scale structured data storage with flexible schemas Provide row level atomic mutation Tables are divided horizontally into tablets One master server and multiple tablet servers Uses Chubby for master election and partial metadata storage Reference: Fay Chang, et. al. Bigtable: A Distributed Storage System for Structured Data. Infrastructure as a Service - Eucalyptus Interfaces compatible with Amazon EC2, S3 and EBS Eucalyputs S3 is used for VM image management Provide virtual network overlay for constructing virtual clusters Reference: Daniel Nurmi, et. al. The Eucalyptus Open-source Cloud-computing System. Dynamic scalability example Elastic Site Dynamic extension of Torque cluster with VMs from clouds
Resource provision based on Job queue status Contextualization completed with Nimbus Context Broker Reference: Paul Marshall, et. al. Elastic Site: Using Clouds to Elastically Extend Site Resources. Comparison of Grids and Clouds Aspect Business model Architecture Resource management Programming model Application model Security Standardization
Grids Collaborative project-oriented Application/Collective/Resource/ Connectivity Batch-scheduled compute model, distributed virtual data model, virtualized workspace and cluster, easy to monitor MPI, Grid RPC, workflow HPC, HTC, scientific gateways, a wide range of applications Security through credential delegations Standardization and interoperability Clouds
Pay per use SaaS/PaaS/IaaS Batch-scheduled as well as interactive compute model, coexistence of centralized and client data model, virtualized hardware and software, hard to monitor MapReduce, declarative programming model, scripting, Web Service Gateways, Web 2.0, SaaS Security through isolation Lack of standards for clouds interoperability Reference: Luis M. Vaquero, et. al. A Break in the Clouds: Towards a Cloud Definition. Reference: Ian Foster, et. al. Cloud Computing and Grid Computing 360-Degree Compared. Database research opportunities related to
Cloud - Revisiting database engines: data intensive applications such as media delivery, peta-scale OLAP systems, power awareness, etc. - Declarative programming for emerging platforms: LINQ, PigLatin, etc. - The interplay of structured and unstructured data: manage collection of structured, semi-structured and unstructured data, context management, etc. - Cloud data services: virtualized database consolidation, better manageability, etc. - Mobile applications and virtual worlds: synthesis of heterogeneous data streams from virtual worlds Reference: Rakesh Agrawal, et. al. The Claremont Report on Database Research. Peta-scale data warehousing at Yahoo! Query cluster
Master cluster Load cluster Everest: a SQL compliant data warehousing engine for analytical applications Built on commodity hardware: k-way mirroring for availability Column based table storage for efficient analytical operations Managing petabytes of data at Yahoo! Reference: Mona Ahuja, et. al. Peta-Scale Data Warehousing at Yahoo!. Database as a Service Compare the performance and scalability of different Database as a Service
implementations from cloud providers with TPC-W Services tested: AWS MySQL, AWS MySQL/R, AWS RDS, AWS SimpleDB, AWS S3, Google AppEngine, Azure SQL Server EB: Emulated browser requests 1EB: ~500 requests/hour 9000EB: ~1250 requests/second Reference: Donald Kossmann, et. al. An Evaluation of Alternative Architectures for Transaction Processing in the Cloud. Parallel Database vs. MapReduce Many MapReduce applications complete data manipulation or search tasks that could be done by parallel databases Compare the performance of parallel databases and MapReduce framework for these data intensive applications Reference: Donald Kossmann, et. al. An Evaluation of Alternative Architectures for Transaction Processing in the Cloud. What I have done
VBS-Lustre: a distributed block storage system for cloud infrastructures Implementation based on the Lustre file system Snapshot s LV1 /lost+found /etc /usr Attachment VM 1 VM 2 .
VBS-Lustre LV2 . Attachment Cloud environment LV: logical volume VM: virtual machine Snapshot: a static copy of a logical volume at a specific time point VBS-Lustre architecture Lustre servers File 1
File 1 Obj n File 2 Obj m VMM Delegate VM VMM Lustre Client Volume Delegate Non-VMM Lustre Client
VBSLustre Service Client : Data transmission : Invocation Preliminary performance test I/O throughput tests done with Bonnie++ Conclusion Cloud is a big step forward based on Grids. Challenges in terms of security, virtualization, QoS, interoperability, etc. Research opportunities for researchers from both distributed systems and database
communities. References  Tran Doan Thanh, et. al. A Taxonomy and Survey on Distributed File Systems, 4th International Conference on Networked Computing and Advanced Information Management.  Frank Schmuck, et. al. GPFS: A Shared-Disk File System for Large Computing Clusters, Proceedings of the FAST 2002 Conference on File and Storage Technologies.  Sun Microsystems, Inc. Lustre File System white paper, 2008.  Klaus Krauter, et. al. A Taxonomy and Survey of Grid Resource Management Systems, SoftwarePractice & Experience, Volume 32, Issue 2 (February 2002).  Ewa Deelman, et. al. Workflows and e-Science: An overview of workflow system features and capabilities, Future Generation Computer Systems, Volume 25, No. 5 (10 May 2009).  Geoffrey Fox, et. al. Workflow in Grid Systems, Concurrency and Computation: Practice & Experience, Volume 18, Issue 10 (August 2006).  Sanjay Ghemawat, et. al. The Google File System, SOSP 2003.  Jeffrey Dean, et. al. MapReduce: Simplified Data Processing on Large Clusters, OSDI 2004.  Fay Chang, et. al. Bigtable: A Distributed Storage System for Structured Data, OSDI 2006.  Luis M. Vaquero, et. al. A Break in the Clouds: Towards a Cloud Definition, ACM SIGCOMM Computer Communication Review,
Volume 39, Number 1, January 2009.  Ian Foster, et. al. Cloud Computing and Grid Computing 360-Degree Compared, GCE 2008.  Daniel Nurmi, et. al. The Eucalyptus Open-source Cloud-computing System, Proceedings of Cloud Computing and Its Applications, October 2008.  Paul Marshall, et. al. Elastic Site: Using Clouds to Elastically Extend Site Resources, CCGrid 2010.  Rakesh Agrawal, et. al. The Claremont Report on Database Research, ACM SIGMOD Record, Volume 37, Issue 3 (September 2008).  Stefan Aulbach, et. al. A Comparison of Flexible Schemas for Software as a Service, SIGMOD 2009.  Andrew Pavlo, et. al. A Comparison of Approaches to Large-Scale Data Analysis, SIGMOD 2009.  Mona Ahuja, et. al. Peta-Scale Data Warehousing at Yahoo!, SIGMOD 2009.  Donald Kossmann, et. al. An Evaluation of Alternative Architectures for Transaction Processing in the Cloud, SIGMOD 2010.  Jinbao Wang, et. al. Indexing Multi-dimensional Data in a Cloud System, SIGMOD 2010.  Xiaoming Gao, et. al. Building a Distributed Block Storage System for Cloud Infrastructure, Proceedings of CloudCom 2010 Conference, IUPUI Conference Center, Indianapolis, November 30-December 3, 2010. Thanks! Distributed file systems
Design principles: - Architecture: centralized vs. cluster based, symmetric vs. asymmetric, etc. - File operation processes: stateful vs. stateless - Communication protocols: RPC/TCP or UDP, InfiniBand, Elan, etc. - Metadata management: central vs. distributed - Synchronization: advisory vs. mandatory locks, segment vs. object locks, etc. - Consistency and replication - Fault tolerance: failure as norm vs. failure as exception - Security: authentication, authorization, privacy Grid resource management systems Design principles: - Machine organization: flat, cells, hierarchical -
Resource model: schema vs. object model, fixed vs. extensible - Resource namespace: relational, hierarchical, hybrid, graph - QoS support: none, soft, hard - Resource information organization: network directory vs. distributed objects - Resource discovery: query (centralized or distributed) based vs. agents
based - Resource dissemination: batch/periodic vs. online/on-demand, push vs. pull - Scheduler organization: centralized, hierarchical, decentralized - State estimation: predictive vs. non-predictive - Scheduling policy: fixed vs. extensible, system oriented vs. application oriented
- Rescheduling: periodic/batch vs. event-driven/online Grid workflow systems Workflow life cycle: - Composition: textual, graphical, compact, semantic - Representation: directed graphs, petri-nets, UML - Execution control models: control flow vs. data flow - Mapping workflow to resources: user-defined, scheduler and broker based, dynamic optimization - Execution: execution models, fault tolerance, adaptive workflow - Provenance: provenance in design stage, provenance for transformed workflow execution - Interoperability An example - Triana:
- Compact graphical composition, Cyclic graph representation, data flow execution model - Scheduler and broker based resource mapping - Job level and service level execution, passive failure report, adaptive workflows - Detailed provenance recording Database as a Service Comparison of different database consolidation schemes Database consolidation: provide virtual databases to multiple tenants with one shared physical database (SQL Server) (SQL Server) (SQL Server)
(DB2) Reference: Stefan Aulbach, et. al. A Comparison of Flexible Schemas for Software as a Service. (HBase) Database as a Service Comparison of different database consolidation schemes Reference: Stefan Aulbach, et. al. A Comparison of Flexible Schemas for Software as a Service. Indexing multi-dimensional data Storage node Storage node Storage
node Storage node C2 overlay network Local R-tree index Storage node C2: a hybrid of CAN and Chord with average hop number of log(N/4)
RT-CAN: a multi-dimensional indexing scheme for both analytical and transactional queries Use a combination of content addressable network (CAN) and R-Tree to index multidimensional data Does not consider dynamic provisioning and management of cloud resources Reference: Mona Ahuja, et. al. Peta-Scale Data Warehousing at Yahoo!.
Dr Nag Diabetes and GM clinic Monday PM Day Primary Care A&E dept Self Referral Community Self Care Intermediate / Community beds Primary care Mental Health Social Care Acute Medical Assessment Unit ITU/HDU Specialist care General care AMBULATORY EMERGENCY CARE...
Theory of Natural Selection. A theory is a unifying idea that explains observations. In over 150 years this is still the best explanation for all the evidences of evolution. Evolution is a slow change over time. It is based on...
Prevention - Education - Treatment - Sustainable Livelihoods . For most of us in the U.S., mosquitoes are merely a nuisance. But in many parts of the world, mosquitoes carry malaria- a preventable, treatable disease that can be deadly if...
Students leave to study, medicine, law, maths, history, media, education, et al at a host of universities including Russell Group universities Oxbridge. Plus Higher Apprenticeships with Boots, Rolls Royce, Balfour Beatty and others. Our students enjoy excellent facilities in a...
Put First Things First Will and Won't Power Packing More into Your Life Better you organize yourself, the more you'll be able to "pack in" Time Quadrants Important Your most important things, your first things, activities that contribute to your...
France - 2.1 Baby Boom! How does the fertility rate of France compare with the average EU rate? Outline how the French attitude to family values differfromthat in eastern Asia. Which EU countries follow the Asian approach? Whatis the linkbetweenfertility...
Special Populations and Support Services - HIV/AIDS, Substance Abuse and Mental Health IUL, in collaboration with the Indiana State Department of Health, Project Renew, Prisoner's Reform United, Marion County Health and Hospital Corporation and Community Centers of Indianapolis (CCI) provides...
Early History of the Atom. Anything with mass and volume. ... "Billiard Ball" Model of the atom. Dalton (1766 -1844) Dalton's "Billiard . Ball" Model. Dalton's Theory. All elements are made up of tiny indivisible particles. Atoms of the same...
Ready to download the document? Go ahead and hit continue!