Download presentation
Presentation is loading. Please wait.
Published byDaisy Stone Modified over 9 years ago
1
Networked Storage Technologies Douglas Thain University of Wisconsin thain@cs.wisc.edu GriPhyN NSF Project Review 29-30 January 2003 Chicago
2
229 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu Unreliable Internet The Problem of Remote I/O Storage Job Remote CPUs Survive disconnections. Hide high latencies. Hide bursty throughput. Audit progressive results. Ensure consistency between job and storage. Arbitrate between users. Make it easy.
3
329 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu NeST Turns Raw Storage into a Storage Appliance Storage NeST Allocable Auditable Authentic Accessible Appl Cmd Tool Web Browser User-Level Adapter Appl OS Kernel Admin or Owner Chirp NFS HTTP GridFTP DaP Match Maker ClassAds
4
429 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu DaP Makes Data Transfer a Managed Job Storage NeST DaP Storage NeST Submit, Query, Remove Data Mvmt Queue AllocationActivation Transfer
5
529 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu Building the Grid Storage Job Remote CPUs NeST DaP GridFTP Transfer Chirp Reservation Adapter Chirp Input Kangaroo Output Chirp Output Status and Supervision Policy and Control Requests
6
629 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu Ph.D. Research Enabled by Griphyn l NeST: Network Storage Technologies –John Bent –http://www.cs.wisc.edu/condor/nest l DaP: Data Placement Manager –Tefvik Kosar –http://www.cs.wisc.edu/condor/dap l Distributed I/O using Grid Services –Douglas Thain –http://www.cs.wisc.edu/~thain l Grid Security –Ian Alderman l Too many MS and BS students to list!
7
729 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu Future Work: l I/O - CPU Specialization in Workloads –Automatically provision a cluster with the correct number of storage/worker nodes. l DaP / DAG integration –Convergence of technologies for reliable data scheduling and reliable job scheduling. l Error Management –What happens when something goes wrong? Backup/retry/pause/inform? l Security –An online CA to issue task-specific certificates just- in-time for work to be done.
8
829 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu Publications Enabled by Griphyn l “Architectural Implications of Pipeline and Batch Sharing in Scientific Workloads”, UW-CS-TR 1463, 2003, also in review. –http://www.cs.wisc.edu/condor/doc/profiling-tr.pdf l “The Case for Sparse Files”, UW-CS-TR 1464, 2003. –http://www.cs.wisc.edu/~thain/library/sparse.pdf l “Error Management in the Pluggable File System”, UW-CS-TR 1448, 2002. –http://www.cs.wisc.edu/condor/doc/pfs-tr.pdf l “Flexibility, Manageability, and Performance in a Grid Storage Appl”, HPDC 2002. –http://www.cs.wisc.edu/condor/nest/papers/nest-hpdc02.pdf l “Error Scope on a Computational Grid: Theory and Practice”, HPDC 2002. –http://www.cs.wisc.edu/condor/doc/error-scope.pdf l “Exploiting Gray-Box Knowledge of Buffer-Cache Management”, USENIX 2002. –http://www.cs.wisc.edu/wind/Publications/dust-usenix02.pdf l “Gathering at the Well: Creating Communities for Grid I/O”, SC 2001. –http://www.cs.wisc.edu/condor/doc/community-sc2001.pdf l “The Kangaroo Approach to Data Movement on the Grid”, HPDC 2001. –http://www.cs.wisc.edu/condor/doc/kangaroo-hpdc10.pdf
9
929 Jan 2003 Douglas Thain, University of Wisconsin thain@cs.wisc.edu The Real Value: “Why don’t you go to down to visit Fermi next Wednesday?”
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.