Cloud Object Storage 101 - Storage Networking Industry Association

Cloud Object Storage 101Live WebcastJuly 14, 201610:00 am PT

Cloud Object Storage 101Live Webcast July 14thNancy BennisDirector, Partner SalesIBM Cloud ObjectStorageAlex McDonaldChair - SNIACloud StorageNetApp2

SNIA Legal Notice!!The material contained in this presentation is copyrighted by the SNIA unlessotherwise noted.Member companies and individual members may use this material inpresentations and literature under the following conditions:!!!!!Any slide or slides used must be reproduced in their entirety without modificationThe SNIA must be acknowledged as the source of any material used in the body of anydocument containing material from these presentations.This presentation is a project of the SNIA Education Committee.Neither the author nor the presenter is an attorney and nothing in thispresentation is intended to be, or should be construed as legal advice or anopinion of counsel. If you need legal advice or a legal opinion please contactyour attorney.The information presented herein represents the author's personal opinion andcurrent understanding of the relevant issues involved. The author, the presenter,and the SNIA do not assume any responsibility or liability for damages arisingout of any reliance on or use of this information.NO WARRANTIES, EXPRESS OR IMPLIED. USE AT YOUR OWN RISK.

AgendaGain understanding of:!!!!!Current storage landscapeObject storage and the movement to cloudBenefits and economics of object storageUse cases and solutionsApplication ecosystem and cloud deployment

Unstructured data growth createdThe need for cloud storage500Projected exabytesKey issues§ Traditional storage designedfor transaction, notunstructured data§ “Growth on growth”- Storing bigger data objects- Storing more data objects- Every new technologycreates new hart source: IDC, March 2014

Data growth is massiveA confluence of forces and sources—including cloud, mobile, social,analytics—create exponential data growth.332%90%of total mobile data trafficwill be cloud apps by 20192growth in mobile data trafficbetween 2015 and 2018110x80%growth of the amount ofdata on the planet by 20203of all data is unstructured (web,social, video, audio, pictures,scans, email)41. Extrapolated from Gartner press release, http://www.gartner.com/newsroom/id/30986172. Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2014-20193. IDC annual Digital Universe study, -takes-off4. IBM data. 2016 IBM CorporationPage6

AgendaGain understanding of:!!!!!Current storage landscapeObject storage and the movement to cloudBenefits and economics of object storageUse cases and solutionsApplication ecosystem and cloud deployment

The evolving storage landscape!Block FileTransactionaldataDatabase dataHigh I/OPerformance1980s UnstructureddataHierarchicalDirectorydefined files1990sObject Large volumesand objectsUnstructureddata filesNo systemoverhead2008

Define object storageEveryone is moving to Cloud Software defined storage (SDS)Low cost server/disk arrays Single virtual pool of storage Distributed access Unprecedented scaleData protection using Erasure coding orRAID Increased data integrity and availability*Object storage refers to a system where data is stored in discrete buckets or“objects,” in contrast to the directories and subdirectories of a traditional file system.By 2019, 30% of midsizeorganizations willleverage public cloudIaaS for backup.5%Up from todayBy 2018, 70% of business andapplication owners will havemore self-service control overtheir data protection services.30%Up from todayBy 2018, the number ofenterprises using the cloudas a backup destination willdouble.11%Up from todaySOURCE: Gartner MagicQuadrant for Data Center Backupand Recovery Software08 June 2016

How Object Storage Technology WorksGeographically Dispersed Erasure CodingDATA1Accessers 2345SITE 1678SITE 291Data is virtualized,encrypted, andsliced usingInformationDispersalAlgorithms.2Slices aredispersed toseparate disks,storage nodesand geographiclocations.10 11 12SITE 3Slicestors 32345678DATAAccessers SCALABLERELIABLESECUREEven withindividual serversor entire sitesdown, real time bitperfect data isretrieved from asubset of slices.

Geographically DispersedErasure CodingSecurityEconomicsLife-Cycle CostLife-Cycle Cost‘05Erasure Coding‘05 Erasure Coding‘08 Commodity H/W‘10 Configurable Efficiency‘14 3.3 Watts per Usable TB‘14 Packed StorageFuture: Automated TieringReliability!‘06 SliceRebuilder‘07 Perfect Bits Data Integrity!‘07 Sixteen 9s!‘09 StrongData ConsistencyGuarantees‘10 Multi-Fault RebuildConfidentiality‘07 KeyInside Threshold Encryption‘13 Secure Multi-tenancy‘13 LDAP‘13 Multi-AD DomainsFuture: ZIG RebuilderCarrier GradeScalability10,000 use casesPerformance‘07 SmartClient‘10 Coast-to-Coast Network‘11 10GB Accessers‘12 Distributed Rebuilder‘13 Multipart Uploads‘14 Cooperative RebuilderFuture: Automated TieringSimplicityAdministration’08 HTTP/RESTManager Interface’11 Zero Downtime Adds,Moves, Changes‘12 Zero DowntimeUpgrades‘12 One Admin per 25PBIntegration‘12 OpenStack Swift‘12 Hadoop HDFS‘12 Partner Solution Network‘12 S3 Interface SupportFuture: IRODSSystem Scale’08 Petabyte Excellence’09 Namespace Registry‘09 No-Database Address’10 Stateless Accesser’10 Distributed Multi-Writer’12 Ten Exabyte SupportFuture: Zettabyte Scale

AgendaGain understanding of:!!!!!Current storage landscapeObject storage and the movement to cloudBenefits and economics of object storageUse cases and solutionsApplication ecosystem and cloud deployment

A highly reliable storage system formassive dataMore efficient utilization of storageRAID 6 Replication!Object Storage!Onsite mirror1.20 PB Raw!Original1.20 PB Raw!!567 TB Raw!!567 TB Raw!567 TB Raw!Remote copy!1.20 PB Raw!1 PBUsable Storage 1 PB3.6 PB9003.6x3.6x3 FTEReplication/backupRaw Storage1.7 PB4TB Disks 432Racks Required 1.7xFloor Space 1.7xOps Staffing.5 FTENone 70% TCO Savings

Object storage economicsObject Storage vs S3 10,000,000Object Storage vs NASS3 1,053dsNet 8,000,000dsNet Object Protected 6,000,000 1,613NAS Gateway and dsNet 4,000,000 4,210 2,000,000Current NAS Single Copy 0480 TB960 TB1920 TB 8,4003840 TBCurrent NAS DR Protected 5 Year TCO comparison Amazon S3 published prices and capacitydiscounts(pricing as of 10/31/14, assumes 20% reads)Cost: 29–61% lower /TB comparison Analysis by IBM Cloud Object StoragecustomerCost: 80% lower

AgendaGain understanding of:!!!!!Current storage landscapeObject storage and the movement to cloudBenefits and economics of object storageUse cases and solutionsApplication ecosystem and cloud deployment

Scalable, shared, secure cloudStorage for the enterpriseBusinessApplicationsExchange EmailSharepointNFS File ServersAudio/VIdeoContentApplicationsData Management SoftwareCloud StorageObjectStorage Unstructured dataNearline performanceNo backup requiredNo limit to scaleNo replicationNo downtimeDR built-in

Enterprise solutions

Use Case: Backup1.10 Data Facility Storage Management SubsystemLeverage low-cost, balanced performance commodity disk storage for weekly/monthly snapshots with higher rates of access requirementsCartridge mediaNFSChallengeSolutionIBM workstationsdfprmm Cost of storingStorage recentdssbackups toohierarchyhigh cost DFSMS componentsproprietary disk storagetvs hsmsystemszSeries ServerNFSUnstructuredDataIBM SpectrumProtectDASD Complexity of backupinfrastructure todayBenefits Secure data at rest withoutreplicationzSeries ServerFigure 1-10 Data Facility Storage Management Subsystem (DFSMS) Scalable backup and alwayson data availability fordependable recovery andsecurityBackupSoftware Expensivedisk to disk to tapeDFSMS componentbackupDFSMS consists of a suite of related data and storage management products for the z/OS Restore and recover fromdisasters faster Time consuming backupPublic orprocessesThe heart of DFSMS is the storage management subsystem (SMS). Using SMS, the storage Efficient distributed backupsacross sites (Cross-Regionofferings)system. DFSMS is an operating environment that helps automate and centralize themanagement of storage based on the policies that your installation defines for availability,performance, space, and security.Private,Hybridadministrator defines policies that automate the management of storage and hardwaredevices. These policies describe data allocation characteristics, performance and availabilitygoals, backup and retention requirements, and storage requirements for the system. Difficult to scale backup withexplosive data growthDFSMS provides a range of automated data and space managementOn-Premisefunctions:! Eliminate,simplify, andautomate tasks normally done by users or a storage administrator How tobackupBYOD,! Improve storage space useObject StorageControl sharingexternal storage centrallyROBO,! file! Enable the storage administrator to manage storage growthDFSMS makes it easier to convert to new device types, and takes advantage of whatCloud Complete data protection forBYOD, ROBO1

Backup: Example business challengeBEFORE: Backup and disaster recoveryLegacy Data SourcesPrimary DataPrimary DataDisk BackupChallengesDRReplicationTape LibraryPrimary DataD/R Time consuming &expensive Redundant data Protecting PB’s Data viacopies, replication &encryption Storage sprawl 18 – 24 month techrefresh Data security Expensive Difficult to manage

Backup: Example solution valueAFTER: Optimized backup and disaster recoveryBenefitsLegacy Data SourcesOperational Gains§ Scalable§ Always ON§ Zero Touch Security§ Easy to ManagePrimary DataDeep archiveFastDiskBackupTape LibraryBACKUP TO CLOUDOn-PremiseCloudObject StorageEconomic Advantage§ Minimize§ Copies/Replication§ Encryption Devices§ Dedupe Hardware§ Simplify backup/restoreprocess§ Reduce storage costs by80%§ De-Duplication Included§ Archive and analyticsready

Use Case: Archive1.10 Data Facility Storage Management SubsystemKeep content accessible with a scalable, reliable and secure long-term dataarchive / data retentionCartridge mediaNFSCHALLENGEdfp Storage & managementcost of rmmStoragedsshierarchyDFSMS componentstraditional tapestorage systems Retention and retrievalmanagementSOLUTIONIBM workstationszSeries ServerNFSUnstructuredDatatvshsmDASD Data retrieval timeArchiveSoftware 80% of DFSMSdatacomponenton Tier 1 storage isDFSMS consistsof a suite90of relateddata and storage management products for the z/OSnot accessedafterdays*system. DFSMS is an operating environment that helps automate and centralize the Rock-solid reliability andavailability provide continuousaccess to data even in the eventof site outagesof storage based on the policies that your installation defines for availability, Meetingmanagementcomplianceperformance,space, and security.Private,Public orrequirementsThe heart of DFSMS is the storage management subsystem (SMS). Using SMS, the storageHybridadministrator defines policies that automate the management of storage and hardwaredevices. These policies describe data allocation characteristics, performance and availabilitygoals, backup and retention requirements, and storage requirements for the system.* IDC ResearchDFSMS provides a range of automated data and space management functions:On-Premise!!!!CloudEliminate, simplify, and automate tasks normally done by users or a storage administratorImprove storage space useObject StorageControl external storage centrallyEnable the storage administrator to manage storage growthDFSMS makes it easier to convert to new device types, and takes advantage of what Keep content accessible inrealtime with a scalable, reliableand secure always on dataarchive Support for all major archivesoftware products directly orthrough gatewayzSeries Server Lost tapes/mediaFigure 1-10 Data Facility Storage Management Subsystem (DFSMS)BENEFITS Keep data on the right tier atthe right time: Lifecycle datamanagement Decrease costs/reclaim of Tier1storage2