Locality Aware dCache & Discussion on Sharing Storage USATLAS Facilities Meeting SMU October 12, 2011.

Slides:



Advertisements
Similar presentations
Status GridKa & ALICE T2 in Germany Kilian Schwarz GSI Darmstadt.
Advertisements

SDN + Storage.
Distributed Xrootd Derek Weitzel & Brian Bockelman.
Ch 7-1 Working with workgroups-1. Objectives Working with workgroups Creating a workgroup Determining whether to use centralized or group sharing.
LANs and WANs Network size, vary from –simple office system (few PCs) to –complex global system(thousands PCs) Distinguish by the distances that the network.
1 Minimal Spanning Tree This is a technique used when we need to connect many places to a system(network).
Large Scale File Distribution Troy Raeder & Tanya Peters.
Chapter 10 Introduction to Wide Area Networks Data Communications and Computer Networks: A Business User’s Approach.
Copyright©2008 N.AlJaffan®KSU1 Chapter 8 Communications and Networks.
Damian Gordon.  When we hook up computers together using data communication facilities, we call this a computer network.
(part 3).  Switches, also known as switching hubs, have become an increasingly important part of our networking today, because when working with hubs,
Site Report: ATLAS Great Lakes Tier-2 HEPiX 2011 Vancouver, Canada October 24 th, 2011.
Outline Network related issues and thinking for FAX Cost among sites, who has problems Analytics of FAX meta data, what are the problems  The main object.
1 Wide Area Network. 2 What is a WAN? A wide area network (WAN ) is a data communications network that covers a relatively broad geographic area and that.
Questionaire answers D. Petravick P. Demar FNAL. 7/14/05 DLP -- GDB2 FNAL/T1 issues In interpreting the T0/T1 document how do the T1s foresee to connect.
Circuit Services - IPTV Christian Todorov Internet2 Fall Member Meeting October 9, 2007.
© 2011 Cisco All rights reserved.Cisco Confidential 1 APP server Client library Memory (Managed Cache) Memory (Managed Cache) Queue to disk Disk NIC Replication.
ATLAS federated xrootd monitoring requirements Rob Gardner July 26, 2012.
Advanced Operating Systems - Spring 2009 Lecture 21 – Monday April 6 st, 2009 Dan C. Marinescu Office: HEC 439 B. Office.
Module – 4 Intelligent storage system
Hadoop Hardware Infrastructure considerations ©2013 OpalSoft Big Data.
Tier 3 Data Management, Tier 3 Rucio Caches Doug Benjamin Duke University.
Oracle Advanced Compression – Reduce Storage, Reduce Costs, Increase Performance Session: S Gregg Christman -- Senior Product Manager Vineet Marwah.
Introduction to dCache Zhenping (Jane) Liu ATLAS Computing Facility, Physics Department Brookhaven National Lab 09/12 – 09/13, 2005 USATLAS Tier-1 & Tier-2.
Grid Lab About the need of 3 Tier storage 5/22/121CHEP 2012, The need of 3 Tier storage Dmitri Ozerov Patrick Fuhrmann CHEP 2012, NYC, May 22, 2012 Grid.
Introduction to DFS. Distributed File Systems A file system whose clients, servers and storage devices are dispersed among the machines of a distributed.
©2015 EarthLink. All rights reserved. Network Diagnostics Professional Services.
AGLT2 Site Report Shawn McKee/University of Michigan HEPiX Spring 2012 Prague, Czech Republic April 23 th, 2012 April 23 th, 2012.
Oracle's Distributed Database Bora Yasa. Definition A Distributed Database is a set of databases stored on multiple computers at different locations and.
Status & Plan of the Xrootd Federation Wei Yang 13/19/12 US ATLAS Computing Facility Meeting at 2012 OSG AHM, University of Nebraska, Lincoln.
Data Access for Analysis Jeff Templon PDP Groep, NIKHEF A. Tsaregorodtsev, F. Carminati, D. Liko, R. Trompert GDB Meeting 8 march 2006.
ATLAS Great Lakes Tier-2 (AGL-Tier2) Shawn McKee (for the AGL Tier2) University of Michigan US ATLAS Tier-2 Meeting at Harvard Boston, MA, August 17 th,
Factors affecting ANALY_MWT2 performance MWT2 team August 28, 2012.
ATLAS XRootd Demonstrator Doug Benjamin Duke University On behalf of ATLAS.
BNL Service Challenge 3 Status Report Xin Zhao, Zhenping Liu, Wensheng Deng, Razvan Popescu, Dantong Yu and Bruce Gibbard USATLAS Computing Facility Brookhaven.
Use Cases for High Bandwidth Query and Control of Core Networks Greg Bernstein, Grotto Networking Young Lee, Huawei draft-bernstein-alto-large-bandwidth-cases-00.txt.
Data Indexing in Peer- to-Peer DHT Networks Garces-Erice, P.A.Felber, E.W.Biersack, G.Urvoy-Keller, K.W.Ross ICDCS 2004.
Simple Infrastructure to Exploit 100G Wide Are Networks for Data-Intensive Science Shawn McKee / University of Michigan Supercomputing 2015 Austin, Texas.
Point-to-point Architecture topics for discussion Remote I/O as a data access scenario Remote I/O is a scenario that, for the first time, puts the WAN.
Data Placement Intro Dirk Duellmann WLCG TEG Workshop Amsterdam 24. Jan 2012.
Storage Classes report GDB Oct Artem Trunov
© 2015 Pittsburgh Supercomputing Center Opening the Black Box Using Web10G to Uncover the Hidden Side of TCP CC PI Meeting Austin, TX September 29, 2015.
HP Openview NNM: Scalability and Distribution. Reference  “HP Openview NNM: A Guide to Scalability and Distribution”,
Efi.uchicago.edu ci.uchicago.edu Storage federations, caches & WMS Rob Gardner Computation and Enrico Fermi Institutes University of Chicago BigPanDA Workshop.
CMS: T1 Disk/Tape separation Nicolò Magini, CERN IT/SDC Oliver Gutsche, FNAL November 11 th 2013.
Strawman LHCONE Point to Point Experiment Plan LHCONE meeting Paris, June 17-18, 2013.
Wide-Area Xrootd testing at MWT2 Sarah Williams Indiana University US ATLAS Tier2/Tier3 Workshop at UChicago
1 SRM v2.2 Discussion of key concepts, methods and behaviour F. Donno CERN 11 February 2008.
LHCONE NETWORK SERVICES: GETTING SDN TO DEV-OPS IN ATLAS Shawn McKee/Univ. of Michigan LHCONE/LHCOPN Meeting, Taipei, Taiwan March 14th, 2016 March 14,
Storage at the ATLAS Great Lakes Tier-2 Tier-2 Storage Administrator Talks Shawn McKee / University of Michigan OSG Storage ForumShawn McKee1.
ATLAS Tier-2 Storage Status AGLT2 OSG Storage Forum – U Chicago – Sep Shawn McKee / University of Michigan OSG Storage ForumShawn McKee1.
LHCONE NETWORK SERVICES INTERFACE (NSI) POINT-TO-POINT TESTBED WITH ATLAS SITES Shawn McKee/Univ. of Michigan Kaushik De/Univ. of Texas Arlington (Thanks.
SLACFederated Storage Workshop Summary Andrew Hanushevsky SLAC National Accelerator Laboratory April 10-11, 2014 SLAC.
VM Technologies (and others) for Site and Service Resiliency Shawn McKee/University of Michigan OSG AHM - Lincoln, Nebraska March 19 th, 2012.
Security recommendations for dCache
AmLight-ExP: Collaboration at 100G
Media Access Methods MAC Functionality CSMA/CA with ACK
Joint AGLT2-MWT2 Networking meeting
Global Data Access – View from the Tier 2
Status of the SRM 2.2 MoU extension
dCache “Intro” a layperson perspective Frank Würthwein UCSD
dCache – protocol developments and plans
Wide Area Network.
ATLAS Sites Jamboree, CERN January, 2017
Web Caching? Web Caching:.
Brookhaven National Laboratory Storage service Group Hironori Ito
DCache things Paul Millar … on behalf of the dCache team.
Internet2 Tech Exchange
An Introduction to Computer Networking
Peta-Cache: Electronics Discussion I Gunther Haller
Presentation transcript:

Locality Aware dCache & Discussion on Sharing Storage USATLAS Facilities Meeting SMU October 12, 2011

dCache and Locality-Awareness Oct  For AGLT2 we have seen significant growth in the amount of storage and compute-power at each site.  We currently have a single 10GE connection used for inter- site transfers and it is becoming strained.  Given 50% of resources at each site, 50% of file access will be on the intersite link  Cost for an additional link is $30K/year + equipment  Could try traffic engineering to utilize the other direction on the MiLR triangle BUT this would compete with WAN use  This got us thinking: we have seen pCache works OK for a single node but the hit rate is relatively small. What if we could “cache” our dCache at each site and have dCache use “local” files? We don’t want to halve our storage though! USATLAS Facilities Meeting

10GE Protected Network for ATLAS  We have two “/23” networks for the AGL-Tier2 but a single domain: aglt2.org  Currently 3 10GE paths to Chicago for AGLT2. Another 10GE DCN path also exists (BW limited)  Our AGLT2 network has three 10GE wavelengths on MiLR in a “triangle”  Loss of any of the 3 waves doesn’t impact connectivity for both sites. VRF to utilize 4 th wave at UM Oct USATLAS Facilities Meeting

dCache Storage Organization/Access  dCache organizes storage by pools. This is the “unit” of storage and maps to a device/partition on a node  Pools may be grouped into pool-groups to organize storage. Typically this is done to group storage by owners/users.  dCache defines links to control how pool groups are accessed. You can view links as a prioritized set of rules defining which pool-groups you might use to determine where you will read or write in dCache. Oct USATLAS Facilities Meeting

dCache Pool at AGLT2 Oct Relatively inexpensive ~$202/TB(useable) Uses resilient cabling (active-active) USATLAS Facilities Meeting dCache Pool

Original AGLT2 dCache Config Oct USATLAS Facilities Meeting

dCache and Locality-Awareness Oct At the WLCG meeting in DESY we worked with Gerd, Tigran and Paul on some dCache issues We came up with a ‘caching’ idea that has some locality awareness It transparently uses pool space for cached replicas Working well! USATLAS Facilities Meeting

Configuration Discussion  The dCache configuration on the previous slide relies on the fact that dCache allows pools to participate in more than one pool-group.  All pools at AGLT2 are members of aglt2Pools and a “local” group like aglt2UMPools or aglt2MSUPools  Reads must come from a local pool. Writes can go to any pool.  Read requests for files that are not on a local pool cause a Pool-to-Pool (P2P) transfer from another pool-group which has the requested file. The remote pool-group is treated analogously to “tape”  The resulting local copy is a “cached” replica. Cached replicas don’t show up as using space in dCache and can use “unused” space.  Cached replicas are cleaned via LRU algorithms when space is needed Oct USATLAS Facilities Meeting

Storage Sharing:Can This Be Extended?  The dCache configuration at AGLT2 allows us to better optimize our storage use for ATLAS while minimizing the required inter-site network traffic.  Could something like this be generalized to allow cross-Tier- 2 transparent sharing of files?  An idea might be to treat other dCache instances as “tape” somehow  Having a single dCache instance spanning two sites would be another way but with obvious issues in implementation/use.  Could dCache utilize Federated Xrootd as a “tape” source?  Other ideas for how to share storage? Between dCache sites? More generally (Xrootd dCache)? Oct USATLAS Facilities Meeting

Regional Sharing?  Rob and I have discussed options to better inter-connect MWT2 and AGLT2.  We are “close” in our WAN peering locations  Invest in a regional “meet-me” switch/device in Chicago?  Advantage is inter-site traffic is isolated from other traffic (uses the meet-me switch)  LHCONE could also be a means of better managing inter-site connectivity  One advantage of regional sharing is a reduced need for storage…can rely upon other sites for some of your storage.  Ideas/discussion about possibilities here? Oct USATLAS Facilities Meeting

Discussion… Discussion? Options? Oct USATLAS Facilities Meeting