ceph on intel: intel storage components, benchmarks, and contributions
TRANSCRIPT
Ceph on IntelIntel Storage Components, Benchmarks,
and ContributionsDaniel Ferber
Open Source Software Defined Storage Technologist,Intel Storage Group
Jim THOmson, intelEnterprise Technologist
is Changing
Source: IDC – The Digital Universe of Opportunities: Rich Data and the Increasing Value of the Internet of Things - April 2014
From now until 2020, the size of the digital universe will about double every two years
Information Growth*2XWhat we do with data is changing, traditional storage infrastructure does not solve tomorrow’s problems
Complexity
Shifting of IT services to cloud computing and next-generation platforms
Cloud
Emergence of flash storage and software-defined environments
New Technologies
The World
Explosion
EVERY MINUTE EVERY DAY*
300 HOURSVIDEO UPLOADEDTO YOUTUBE
51,000APPS DOWNLOADED
204MILLION E-MAILS
Source: TechSpartan.co.uk - 2013 vs 2015 In an Internet minute
48 HOURSVIDEO UPLOADEDTO YOUTUBE
47,000APPS DOWNLOADED
200MILLION E-MAILS
2013 2015
Inform a tion
The Impact ofthe Cloud
Empowerment of the end-user through cloud services
Emergence of new technologies and architectures
Shifting the role of information technology professionals
Price + Performance= AffordableFlashReliability and DurabilityStorage
Storage CompressionEncryptionErasure codingDeduplicationFlash (Non-volatile memory)Data tiering
WorkloadOptimization
ArchitecturesStorage SCALE-UP
SCALE-OUT
marketDynamics
STORAGETRANSFORMINGKEY DRIVERS
CLOUDCLOUD SERVICE PROVIDERS1
STORAGE MEDIATRANSITIONS3 3D XPOINT™
TECHNOLOGY
NEXT GENERATION ARCHITECTURES 2
Where Data is Created
0
10
20
30
40
50
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
(ZB)
Unstructured Data Structured Data
Sources: IDC, 2011 Worldwide Enterprise Storage Systems 2011-2015 Forecast updateIDC, The Digital Universe Study In 2020 Forecasts
4.4ZB
44ZB 90%
10% 0%
20%
40%
60%
80%
100%
2012 2013 2014 2015 2016 2017 2018 2019 2020
% of Total Digital Universe
Emerging Markets Mature Markets
Sources: IDC,’s Digital Universe Study, 2014
USA, Canada, Western Europe, Australia, NZ, and Japan
China, India, Mexico, Brazil, and Russia
By 2020, about 90% of all data will beunstructured, driven by Consumer Images,
Voice, and the Web
Emerging Markets will Surpass Mature Markets before 2017 regarding data creation.
Data Creation by Type (ZB)
11
Annual Growth Rate for Data Storage Capacity
10%
25%
15%13%
12%
8%6%
3% 3% 3%
0%2%
0%
5%
10%
15%
20%
25%
30%
1% to 10% annually
11% to 20%
annually
21% to 30%
annually
31% to 40%
annually
41% to 50%
annually
51% to 60%
annually
61% to 70%
annually
71% to 80%
annually
81% to 90%
annually
91% to 100%
annually
More than 100%
annually
Don't know
Source: ESG Research Report: 2015 Data Storage Market Trends
SSD Use in Servers or External Storage Systems
Source: ESG Research Report: 2015 Data Storage Market Trends
14
IntelStorage
Processor D
Delivering the performance and advanced intelligence of Intel® Xeon® processors to dense and low power storage designs
Intel® Xeon®
Latency: ~100xSize of Data: ~1,000x 1000x faster than NAND
1000x higher endurance of NAND10x denser than DRAM
Technology claims are based on comparisons of latency, density, and write cycling metrics amongst memory technologies recorded on published specifications of in-market memory products against internal Intel specifications.
3D XpOINT™TECHNOLOGY
New Class of Non-Volatile Memory
17
Intel® Solid State Drives
“The only SSDs that never ever gave me any issues like timeouts, task aborts… are Intel DC S3700s”
From a post on ceph-devel*
Source: http://ceph.com/r esources/maili ng-‐‑list-‐‑irc
18
Intel’s role in storageAdvance the Industry
Open Source & StandardsBuild an Open Ecosystem
Intel® Storage BuildersEnd user solutions
Cloud, Enterprise
Intel Technology LeadershipStorage Optimized CPU’s
Intel® Xeon® E5v4 2600 PlatformIntel® Xeon® Processor D-1500 Platform
Storage Optimized SoftwareIntel® Intelligent Acceleration Library
Intel® Storage Performance Development Kit
Non-Volatile Memory3D Xpoint™
Intel® Solid State Drives for Datacenter
>7 Cloud storage solutions architectures
70+ partners
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products.
helping customers to enable Next gen storage
Next gen solutions architectures>26
>10 Enterprise storage solution architectures
20
Intel Ceph Contribution Timeline
2014 2015 2016
* Right Edge of box indicates approximate release date
New Key/Value Store Backend (rocksdb)
Giant* Hammer Infernalis Jewel
CRUSH Placement Algorithm improvements(straw2 bucket type)
Bluestore Backend Optimizations for NVM
Bluestore SPDKOptimizations
RADOS I/O Hinting(35% better EC Write erformance)
Cache-tiering with SSDs(Write support)
PMStore(NVM-optimized backend
based on libpmem)
RGW, BluestoreCompression, Encryption
(w/ ISA-L, QAT backend)
Virtual Storage Manager (VSM) Open Sourced
CeTuneOpen Sourced
Erasure Coding support with ISA-L
Cache-tiering with SSDs(Read support)
Client-side Block Cache (librbd)
Ceph @ Intel – 2016 Ceph Focus Areas
21
Optimize for Intel® platforms, flash and networking• Compression, Encryption hardware offloads (QAT & SOCs) • PMStore (for 3D XPoint DIMMs)• RBD caching and Cache tiering with NVM• IA optimized storage libraries to reduce latency (ISA-L, SPDK)
Performance profiling, analysis and community contributions• All flash workload profiling and latency analysis• Streaming, Database and Analytics workload driven optimizations
Ceph enterprise usages and hardening• Manageability (Virtual Storage Manager)• Multi Data Center clustering (e.g., async mirroring)
End Customer POCs with focus on broad industry influence• CDN, Cloud DVR, Video Surveillance, Ceph Cloud Services, AnalyticsPOCs
Ready to use IA, Intel NVM optimized systems & solutions from OEMs & ISVs• Ready to use IA, Intel NVM optimized systems & solutions from OEMs & ISVs• Intel system configurations, white papers, case studies • Industry events coverage
Go to Market
Intel® Storage Acceleration Library (Intel® ISA-L)
Intel® Storage Performance Development Kit (SPDK)
Intel® Cache Acceleration Software (Intel® CAS)
Virtual Storage Manager Ce-Tune Ceph Profiler
22
Ceph Storage Cluster
Hardware Environment Overview
Ceph network (192.168.142.0/24) - 10Gbps
CBT / Zabbix / Monitoring FIO RBD Client
• OSD System Config: Intel Xeon E5-2699 v3 2x@ 2.30 GHz, 72 cores w/ HT, 96GB, Cache 46080KB, 128GB DDR4• Each system with 4x P3700 800GB NVMe, partitioned into 4 OSD’s each, 16 OSD’s total per node
• FIO Client Systems: Intel Xeon E5-2699 v3 2x@ 2.30 GHz, 72 cores w/ HT, 96GB, Cache 46080KB, 128GB DDR4• Ceph v0.94.3 Hammer Release, CentOS 7.1, 3.10-229 Kernel, Linked with JEMalloc 3.6
• CBT used for testing and data acquisition • Single 10GbE network for client & replication data transfer, Replication factor 2
FIO RBD ClientFIO RBD Client
FIO RBD Client
FIO RBD ClientFIO RBD Client
FatTwin (4x dual-socket XeonE5 v3)
FatTwin (4x dual-socket XeonE5 v3)
Ceph OSD
1
NVMe1 NVMe3
NVMe2 NVMe4
Ceph OSD
2
Ceph OSD
3
Ceph OSD
4
Ceph OSD
16
…
Ceph OSD
1
NVMe1 NVMe3
NVMe2 NVMe4
Ceph OSD
2
Ceph OSD
3
Ceph OSD
4
Ceph OSD
16
…
Ceph OSD
1
NVMe1 NVMe3
NVMe2 NVMe4
Ceph OSD
2
Ceph OSD
3
Ceph OSD
4
Ceph OSD
16
…Ceph O
SD1
NVMe1 NVMe3
NVMe2 NVMe4
Ceph OSD
2
Ceph OSD
3
Ceph OSD
4
Ceph OSD
16
…
Ceph OSD
1
NVMe1 NVMe3
NVMe2 NVMe4
Ceph OSD
2
Ceph OSD
3
Ceph OSD
4
Ceph OSD
16
…
SuperMicro 1028U SuperMicro 1028U SuperMicro 1028U SuperMicro 1028U SuperMicro 1028U
Intel Xeon E5 v3 18 Core CPUsIntel P3700 NVMe PCI-e Flash
Easily serviceable NVMe Drives
23
4K Random Read & Write Performance SummaryFirst Ceph cluster to break 1 Million 4K random IOPS
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Any difference in system hardware or software design or configuration may affect actual performance. See configuration slides in backup for details on software configuration and test benchmark parameters.
Workload Pattern Max IOPS
4K 100% Random Reads (2TB Dataset)1.35Million
4K 100% Random Reads (4.8TB Dataset)1.15Million
4K 100% Random Writes (4.8TB Dataset)200K
4K 70%/30% Read/Write OLTP Mix (4.8TB Dataset) 452K
Source: Openstack Summit 2015: Accelerating Cassandra workloads on ceph with all flash pcie ssds
Red Hat Ceph Reference Architecture Documents
24
Meta Formula for Ceph Deployments• Have a general understanding of the use cases you want to support with Ceph
• Understand the kind of performance or cost/performance you want to deliver
• Refer to a reference architecture resource to match your use case(s) with known and measured reference architectures:• http://www.redhat.com/en/resources/performance-and-sizing-guide-red-hat-
ceph-storage-qct-servers
• https://www.redhat.com/en/files/resources/en-rhst-cephstorage-supermicro-INC0270868_v2_0715.pdf
• These documents have Ceph config, tuning and best practices guidance
• Additional help is available from Red Hat, including support and quick start
25
IntelExperiencethe Storage Revolution
in accelerating next generation storage solutions
Flash technology for high performance
to accelerate software-defined storage
Invest
Embrace
Align
Legal noticesCopyright © 2016 Intel Corporation.
All rights reserved. Intel, the Intel logo, Xeon, Intel Inside, and 3D XPoint are trademarks of Intel Corporation in the U.S. and/or other countries.
*Other names and brands may be claimed as the property of others.
FTC Optimization NoticeIntel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice. Notice revision #20110804
The cost reduction scenarios described in this document are intended to enable you to get a better understanding of how the purchase of a given Intel product, combined with a number of situation-specific variables, might affect your future cost and savings. Nothing in this document should be interpreted as either a promise of or contract for a given level of costs.
Backup
30