Good morning Bruce Rosen, chair of the BIRN Executive Committee.
Mouse BIRN - Jagadeeswaran Rajendiran Pipeline Processing Environment.
-
Upload
melanie-gray -
Category
Documents
-
view
216 -
download
0
Transcript of Mouse BIRN - Jagadeeswaran Rajendiran Pipeline Processing Environment.
Mouse BIRNMouse BIRN
- Jagadeeswaran Rajendiran- Jagadeeswaran Rajendiran
Pipeline Processing Pipeline Processing EnvironmentEnvironment
Motivation …Motivation …
Algorithms that have been implementedAlgorithms that have been implemented• for different computer architecturesfor different computer architectures• with varying input/ output needswith varying input/ output needs
Data characteristicsData characteristics• high volumeshigh volumes• located on different machineslocated on different machines• have different formatshave different formats
Computational needsComputational needs• Complex preprocessing/ analysis tasksComplex preprocessing/ analysis tasks• Single, flexible all-encompassing frameworkSingle, flexible all-encompassing framework• High volume of measurements/ resultsHigh volume of measurements/ results
It would be nice to auto-store intermediate results!It would be nice to auto-store intermediate results!
Motivation Motivation
Researchers are expected to have:Researchers are expected to have:• Expert knowledgeExpert knowledge
ToolsTools• Access to computing powerAccess to computing power
Collaboration is difficultCollaboration is difficult• data and analyses data and analyses
(communication)(communication) Seamless translation between Seamless translation between
data formatsdata formats Visualization frameworkVisualization framework
• PortablePortable• Context sensitiveContext sensitive
LONI Pipeline LONI Pipeline
Processing EnvironmentProcessing Environment• data flow execution environmentdata flow execution environment• links independent programslinks independent programs• auto-parallelizes sectionsauto-parallelizes sections• repeatable on different data-setsrepeatable on different data-sets• Uses:Uses:
Integrate your program as part of an Integrate your program as part of an analytic processanalytic process
Send a collaborator an analytic processSend a collaborator an analytic process Validate data formatsValidate data formats Parallel batch processing with error Parallel batch processing with error
handling handling
ArchitectureArchitecture
ModularModular Generic domain of applicationsGeneric domain of applications Platform-independent (Java)Platform-independent (Java) Extensible and flexible designExtensible and flexible design Clean code base Clean code base
• No unnecessary dependenciesNo unnecessary dependencies Rapid prototyping of new Rapid prototyping of new
functionalitiesfunctionalities Single encompassing GUISingle encompassing GUI
• User friendlyUser friendly• Data centricData centric
Feature rich GUIFeature rich GUI
• Zoom• Pan• Connections• Tooltips• Module Lists• Runtime updates• Workspaces
Execution EnvironmentExecution Environment Executes modulesExecutes modules ParallelizationParallelization
• Data-independent sectionsData-independent sections• If inputs are active and readyIf inputs are active and ready
Data-centricData-centric• Intermediate data-setsIntermediate data-sets• Can save for later analysis: OntologyCan save for later analysis: Ontology
Remote ExecutionRemote Execution• Client-Server using Java RMIClient-Server using Java RMI
Seamless integration (external Seamless integration (external apps)apps)
Data VisualizationData Visualization
Additional functionality
Is integrated via the extension
architecture.
• Viewer Extensions• Configurable
Data MediationData Mediation
• Seamless• Format conversion
Grid Engine IntegrationGrid Engine Integration
Horizontal vs. Vertical Parallelization
• VERTICAL— Pipeline
• HORIZONTAL— Grid
BREAK FOR DEMO!BREAK FOR DEMO!
Protocols and ApplicationsProtocols and Applications Data/ Functionality specificationData/ Functionality specification
• XMLXML Network protocolsNetwork protocols
• HTTP, FTPHTTP, FTP• Grid engine capabilitiesGrid engine capabilities
SUN Grid systemSUN Grid system SRB @ UCSDSRB @ UCSD
SecuritySecurity• Client - ServerClient - Server
Diffie Hellman key agreementDiffie Hellman key agreement Advanced Encryption Standard (AES)Advanced Encryption Standard (AES)
• User AuthenticationUser Authentication SHA-1 digest of passphraseSHA-1 digest of passphrase AES key from digestAES key from digest
ApplicationsApplications• SlicerSlicer• ShivaShiva
Upcoming features …Upcoming features …
Improved usabilityImproved usability Fault tolerance (system & Fault tolerance (system &
network)network) EfficiencyEfficiency
• Overlay grid implementationOverlay grid implementation ProvenanceProvenance
• Data validationData validation• Information recoveryInformation recovery
ResourcesResources
Wiki-PageWiki-Page• http://www.http://www.loniloni..uclaucla..edu/twiki/bin/view/Pipeline/edu/twiki/bin/view/Pipeline/
PersonnelPersonnel• Dr. Arthur Toga: Dr. Arthur Toga: [email protected]@loni.ucla.edu• Michael Pan: Michael Pan: [email protected]@loni.ucla.edu• Ron Nor: Ron Nor: [email protected]@loni.ucla.edu• Jags: Jags: [email protected]@loni.ucla.edu
THANK YOU