DSS Data & Storage Services CERN Lustre Evaluation and Storage - - PowerPoint PPT Presentation

dss
SMART_READER_LITE
LIVE PREVIEW

DSS Data & Storage Services CERN Lustre Evaluation and Storage - - PowerPoint PPT Presentation

DSS Data & Storage Services CERN Lustre Evaluation and Storage Outlook Tim Bell Arne Wiebalck HEPiX, Lisbon 20 th April 2010 CERN IT Department CH-1211 Genve 23 Switzerland www.cern.ch/i t DSS Agenda Lustre Evaluation Summary


slide-1
SLIDE 1

Data & Storage Services

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

DSS

CERN Lustre Evaluation and Storage Outlook

Tim Bell Arne Wiebalck HEPiX, Lisbon 20th April 2010

slide-2
SLIDE 2

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Agenda

  • Lustre Evaluation Summary
  • Storage Outlook

– Life cycle management – Large disk archive

  • Conclusions

CERN Lustre Evaluation and Storage Outlook - 2

slide-3
SLIDE 3

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Lustre Evaluation Scope

  • HSM System

– CERN Advanced STORage Manager (CASTOR) – 23 PB, 120 million files, 1’352 servers

  • Analysis Space

– Analysis of the experiments’ data – 1 PB access with XRootD

  • Project Space

– >150 projects – Experiments’ code (build infrastructure) – CVS/SVN, Indico, Twiki, …

  • User home directories

– 20’000 users on AFS – 50’000 volumes, 25 TB, 1.5 billion acc/day, 50 servers – 400 million files

3

slide-4
SLIDE 4

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Evaluation Criteria

  • Mandatory is support for ...

– Life Cycle Management – Backup – Strong Authentication – Fault-tolerance – Acceptable performance for small files and random I/O – HSM interface

  • Desirable is support for ...

– Replication – Privilege delegation – WAN access – Strong administrative control

  • Performance was explicitly excluded

– See the results of the HEPiX FSWG

4

slide-5
SLIDE 5

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Compliance (1/3)

  • Life cycle management

– Not OK: no support for live data migration, Lustre or kernel upgrades, monitoring, version compatibility

  • Backup

– OK: LVM snaphots for MDS plus TSM for files worked w/o problems

  • Strong Authentication

– Almost OK: Incomplete code in v2.0, full implementation expected Q4/2010 or Q1/2011

5

slide-6
SLIDE 6

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Compliance (2/3)

  • Fault-tolerance

– OK: MDS and OSS failover (we used a fully redundant multipath iSCSI setup)

  • Small files

– Almost OK: Problems when mixing small and big files (striping)

  • HSM interface

– Not OK: Not supported yet, but under active development

6

slide-7
SLIDE 7

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Compliance (3/3)

  • Replication

– Not OK: not supported (would help with data migration and availability)

  • Privilege delegation

– Not OK: not supported

  • WAN access

– Not OK: may become possible once Kerberos is fully implemented (cross-realm setups)

  • Strong administrative control

– Not OK: pools not mandatory, striping settings cannot be enforced

7

slide-8
SLIDE 8

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Additional thoughts

  • Lustre comes with (too) strong client/server

coupling

– Recovery case

  • Moving targets on the roadmap

– Some of the requested features are on the roadmap since years, some are simply dropped

  • Lustre aims at extreme HPC rather then a general

purpose file system

– Most of our requested features are not needed in the primary customers‘ environment

8

slide-9
SLIDE 9

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Lustre Evaluation Conclusion

  • Operational deficiencies do not allow for a Lustre-

based storage consolidation at CERN

  • Lustre still interesting for the analysis use case (but
  • perational issues should be kept in mind here as well)
  • Many interesting and desired features (still) on the

roadmap, so it’s worthwhile to keep an eye on it

  • For details, see write up at

https://twiki.cern.ch/twiki/pub/DSSGroup/LustreEvaluation/CERN_Lustre_Evaluation.pdf

9

slide-10
SLIDE 10

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Agenda

  • Lustre Evaluation Summary
  • Storage Outlook

– Life cycle management – Large disk archive

  • Conclusions

CERN Lustre Evaluation and Storage Outlook - 10

slide-11
SLIDE 11

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Life Cycle Management

  • Archive sizes continue to grow

– 28 PB tape used currently at CERN – 20 PB/year expected

  • Media refresh every 2-3 years

– Warranty expiry on disk servers – Tape drive repacking to new densities

  • Time taken is related to

– TOTAL space, not new data volume recorded – Interconnect between source and target – Metadata handling overheads per file

  • Must be performed during online periods

– Conflicts between user data serving and refresh

Storage Outlook - 11

slide-12
SLIDE 12

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Repack Campaign

Storage Outlook - 12

5000 10000 15000 20000 25000 30000 35000 Nov-08 Dec-08 Jan-09 Feb-09 Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Tapes Repacked

Total Tapes Repacked

  • Last repack campaign took 12 months to copy 15PB of data
  • When next drives are available, there will be around 35PB of data
  • To complete repack in 1 year, data refresh will require as much

resources as LHC data recording

  • This I/O capacity needs to be reserved in the disk and tape

planning for sites with large archives

slide-13
SLIDE 13

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Disk Based Archive?

Storage Outlook - 13

2 4 6 8 10 12 14 16 2009 2010 2011 2012 2013 2014 2015 Maximum capacity (TB)

Disk and Tape Capacity Projections

Pessimistic disk projection Optimistic disk projection Pessimistic tape projection Optimistic tape projection

  • Can we build a disk based archive at reasonable cost compared to

a tape based solution?

slide-14
SLIDE 14

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Storage in a Rack

  • Tape Storage at CERN

– 1 drive has 374 TB storage – Average rate 25 MB/s

  • Disk Server equivalent

– 2 head nodes

  • 2 x 4 port SAS cards

– 8 JBOD expansion units

  • 45 x 2 TB disks each

– Capacities

  • 720 TB per rack
  • 540 TB when RAID-6 of 8 disks
  • 270 TB per head node

Storage Outlook - 14

slide-15
SLIDE 15

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS High Availability

Storage Outlook - 15

Head Node Disk Array Disk Array Disk Array Disk Array Disk Array Disk Array Disk Array Disk Array Head Node

slide-16
SLIDE 16

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Simulation 20 PB/yr 2011-15

  • Costs normalised to tape HSM as 100
  • Storage in a rack can be comparable with

tape on cost/GB

Storage Outlook - 16

20 40 60 80 100 120 140 160 180 Tape with just in time repack Storage in a box Storage in a rack Storage in a rack with backup Normalised Cost (Current=100)

5 year archive cost

slide-17
SLIDE 17

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Simulation for power

  • Additional power consumption of 100 kWatt
  • Cost included in the simulation

Storage Outlook - 17

20 40 60 80 100 120 140 160 2011 2012 2013 2014 2015 kWatts

Annual Power Consumption

Nearline Repack Just in Time Storage in a rack Storage in a rack with backup

slide-18
SLIDE 18

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Areas to investigate

  • Reliability

– Corruptions – Scrubbing

  • Availability

– Fail-over testing

  • Power conservation

– Disk spin down / up

  • Lifecycle management

– 40 days to drain at gigabit ethernet speeds

  • Manageability

– Monitoring, Repair, Install

  • Operations cost

– How much effort is it to run

Storage Outlook - 18

slide-19
SLIDE 19

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Conclusions

  • Lustre should continue to be watched but

currently is not being considered for Tier-0, Analysis or AFS replacement

  • Lifecycle management is a major concern

for the future as the size of the archive grows

  • Disk based archiving may be an option

– In-depth reliability study before production – Watch trends for disk/tape capacities and pricing – Adapt software for multiple hierarchies

CERN Lustre Evaluation and Storage Outlook - 19

slide-20
SLIDE 20

Data & Storage Services

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

DSS

Backup Slides

slide-21
SLIDE 21

CERN IT Department CH-1211 Genève 23 Switzerland

www.cern.ch/it

Internet Services

DSS Use Cases

Presentation Title - 21