1 / 22

gLite, the middleware for EGEE – status and future

gLite, the middleware for EGEE – status and future. Bob Jones ISCG 2006, Taipei, Taiwan. www.glite.org. Grid middleware. The Grid relies on advanced software, called middleware , which interfaces between resources and the applications The GRID middleware:

Download Presentation

gLite, the middleware for EGEE – status and future

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. gLite, the middleware for EGEE – status and future Bob Jones ISCG 2006, Taipei, Taiwan

  2. www.glite.org Bob Jones, ISCG 2006

  3. Grid middleware • The Grid relies on advanced software, called middleware, which interfaces between resources and the applications • The GRID middleware: • Finds convenient places for the application to be run • Optimises use of resources • Organises efficient access to data • Deals with authentication to the different sites that are used • Runs the job & monitors progress • Recovers from problems • Transfers the result back to the user Bob Jones, ISCG 2006

  4. The Grid Software Stack Applications APPLICATIONS Grid Application Layer Collective Services Grid MIDDLEWARE Underlying Grid Services GLOBUSCondor (VDT) Infrastructure Fabric Fabric services Bob Jones, ISCG 2006

  5. File and ReplicaCatalogs User Interface Resource Broker Computing Element Storage Element Site X Grid Topology and Services Information System submit query retrieve update credential publish state submit query retrieve AuthorizationService Bob Jones, ISCG 2006

  6. Grid Middleware Key success factors for production quality software: • Strict software process • Use industry standard software engineering methods • Software configuration management, version control, defect tracking, automatic build system, … • Conservative in what software to use • Avoid “cutting-edge” software • Deployment on over 100 sites cannot assume a homogenous environment – middleware needs to work with many underlying software flavors • Avoid evolving standards • Evolving standards change quickly (and sometime significantly cf. OGSI vs. WSRF) – impossible to keep pace on over 100 sites You will not develop and deploy your PhD project on a production Grid infrastructure There is a long (and tedious) path from prototypes to production Bob Jones, ISCG 2006

  7. EGEE Middleware: gLite • gLite • Exploit experience and existing components from VDT (Condor, Globus), EDG/LCG, AliEn, and others • Develop a lightweight stack of generic middleware useful to EGEE applications (HEP and Biomedics are pilot applications). • Should eventually deploy dynamically (e.g. as a globus job) • Pluggable components – cater for different implementations • Focus is on re-engineering and hardening Bob Jones, ISCG 2006

  8. EGEE Middleware • Requirements through database (>400 requirements) and task forces • Applications involved in all stages • Evaluate service functionality on prototypes • Evaluate service reliability and performance on pre-production service • Use the services on the infrastructure • Technical Coordination Group (TCG) • Started in November 2005 • Brings together all technical activities • Direct US participation in middleware design team (Condor and Globus) • Middleware Security Group • Brings together EGEE and other projects • Consolidate middleware release • Current LCG-2.7 and gLite 1.5 distributions will be consolidated and released as gLite 3.0 • Using VDT as common basis for EGEE and OSG is very important for interoperability • EGEE is not only user but also contributor • ETICS project started Jan 1st with the goal to have a common build and test infrastructure • Univ. of Wisconsin is a partner Bob Jones, ISCG 2006

  9. Software stack and origin of servicesin release 1 (simplified) • Storage Element • gLite-I/O (AliEn) • Reliable File Transfer (EGEE) • GridFTP (Globus) • SRM: Castor (CERN), dCache (FNAL, DESY), other SRMs • Catalog • File/Replica & Metadata Catalogs (EGEE) • Security • GSI (Globus) • VOMS (DataTAG/EDG) • Authentication for C and Java based (web) services (EDG) • Computing Element • Gatekeeper (Globus) • Condor-C (Condor) • CE Monitor (EGEE) • Local batch system (PBS, LSF, Condor) • Workload Management • WMS (EDG) • Logging and bookkeeping (EDG) • Condor-C (Condor) • Information and Monitoring • R-GMA (EDG) Now doing rigorous scalability and performance tests on pre-production service Bob Jones, ISCG 2006

  10. Authentication Use of GSI, X.509 certificates Generally issued by national certification authorities Agreed network of trust: International Grid Trust Federation (IGTF) EUGridPMA APGridPMA TAGPMA All EGEE sites will usually trust all IGTF root CAs Authorization Until LCG-2.7.0 via grid-map files only From LCG-2.7.0 using VOMS extended proxies Call-outs to local authorization services Integration with grid services under way – compute elements, storage systems For some time the authorization will be a mixture of call-outs and grid-map files until all services understand extended proxies CAs, Authentication, Authorization APGridPMA EUGridPMA TAGPMA Asia-Pacific Grid PMA The Americas Grid PMA European Grid PMA Bob Jones, ISCG 2006

  11. Job Management: Workload Management – Resource Broker DLI/SI interface to catalogues for data-based scheduling Bulk job submission (gLite-3.0) DAGs (gLite-3.0) Push/pull mode (pull untested – gLite-3.0) Compute Element (CE): Globus/EDG/LCG  Condor_C (VO-based scheduling) in gLite-3.0 Logging & Bookkeeping Local Batch systems: LSF, PBS, Condor, (Sun Grid Engine) Additional tools: Ability to “peek” at stdout/stderr of running jobs User job monitoring – look at the status (state, cpu time, etc) of running jobs Data Management File and replica catalogues (LFC) Central or local (not distributed) Replication via Oracle, or squid caches tested by LCG Secure File Transfer Service (FTS) Reliable data transfer Uses gridftp or srmcopy as transport Storage Elements based on SRM interface DPM: implements Posix ACLs, VOMS roles/groups (gLite-3.0) Other available SEs: dCache, Castor Deprecated: “Classic SE” – basically just gridftp Metadata catalogue: AMGA (gLite-3.0 – partial support) Secure Keystore: Hydra (gLite-3.0 – partial support) Utilities and IO libraries: Lcg-utils GFAL – this is the SRM client library gLiteIO – expect functionality to be replaced Basic Services Bob Jones, ISCG 2006

  12. Functional Tests gLite Software Process Development Integration Testing Deployment Packages Software Code Fail Pass Testbed Deployment Integration Tests Fix Fail Pass Installation Guide, Release Notes, etc Bob Jones, ISCG 2006

  13. gLite Key Concepts • Centered around VOs • It’s ultimately the VO who gets resources allocated and need to decide how to best use them (share them among the VO users) • Distinguish between infrastructure and VO services • Infrastructure services • Operated and trusted by the resource administrator • Implement site policies • Including what share of the resources are allocated to a VO • Provide the required security, auditing, and accounting • Grid and standard services • E.g. batch system, gatekeeper, gridFTP, … Bob Jones, ISCG 2006

  14. gLite Key Concepts • VO services • Implement intra-VO policies • Scheduling, priorities, etc. • Managed and operated by a VO • Typically by sites on behalf of VOs • A service instance may serve multiple VOs • Currently mostly higher level services • Resource brokers, catalogs, … • There is the need of deploying VO services closer to the resource • Better information about the resource and better control about the resource • Downside: more and more services to be deployed at the sites – see discussion later on Bob Jones, ISCG 2006

  15. Guiding Principles Service Oriented Architecture Interoperability Portability Building on existingcomponents in alightweight manner Web Services Modularity AliEn LCG Condor Scalability Globus SRM ... Bob Jones, ISCG 2006

  16. Middleware Challenges: Security • In principle, Grid security requirements are not different from standard security requirements • Users want their data and application secured (including data transfer) • Sites want access to their resources secured and audited • What makes it challenging are the different administrative domains interconnected on the Grid and the need to establish mutual trust Bob Jones, ISCG 2006

  17. Security: Basic Concepts • Grid security is based on X.509 PKI infrastructure • Certificate Authorities (CA) issue (long lived) certificates identifying individuals (much like a passport) • Trust between CAs and sites is established (offline) • User identification is done by using (short lived) proxies of their certificates • Proxies can • Be delegated to a service such that it can act on the user’s behalf • Include additional attributes (like VO information via the VO Membership Service VOMS) • Be stored in an external proxy store (myProxy) • Be renewed (in case they are about to expire) • Standard TLS (like ssh) and MLS (like WS-Sec) is used • MLS has performance and support issues! Bob Jones, ISCG 2006

  18. Security Architecture Bob Jones, ISCG 2006

  19. Middleware Challenges: Data Management • Need common interface to storage resources • Storage Resource Manager (SRM) • Need to keep track where data is stored • File and Replica Catalogs • Need scheduled, reliable file transfer • File transfer and placement services • Need a common security model • ACLs enforcement based on Grid identities – DNs • Heterogeneity • Data is stored on different storage systems using different access technologies • Distribution • Data is stored in different locations – in most cases there is no shared file system or common namespace • Data needs to be moved between different locations • Different Administrative Domains • Data is stored at places you would normally have no access to • Security and auditing implications Bob Jones, ISCG 2006

  20. Fireman File andReplica Catalog Database File Transfer andPlacement Service StorageIndex Transfer Agent FPS Database FTS Data Management Architecture Storage Element WSDL VOMS Storage API Getcredential File I/O gLite I/O SRM gridFTP File namespace and Metadata mgmt Storecredential File replication Proxy renewal ReplicaLocation MyProxy WMS Bob Jones, ISCG 2006

  21. Middleware Challenges: Workload Management • Computational tasks of thousands of users need to be scheduled on the available Grid resources • Grid (Meta)Scheduling consists of: • Resource Discovery/Brokering • Find suitable resources • Matchmaking • Assign a job to a resource that satisfies job requirements • Job execution • Reliably execute the jobs and retrieve output • Deal with error management • Job execution requires to find the “right” Computing Element (computing resource) • with maybe boundary conditions (architecture, software installed, data accessible, etc.) Bob Jones, ISCG 2006

  22. gLite Documentationand Information sources • Installation Guide • Release Notes • General • Individual Components • User Manuals • With Quick Guide sections • CLI Man pages • API’s and WSDL • Beginners Guide and Sample Code • Bug Tracking System • Mailing Lists • gLite-discuss • Pre-Production Service • Other • Data Management (FTS) Wiki • Pre-Production Services Wiki • Public and Private • Presentations Bob Jones, ISCG 2006

More Related