Unifying Data and Optimizing Application Performance with Flash

Unifying Data and Optimizing Application Performance with Flash
SOLUTION BRIEF
UNIFYING DATA AND OPTIMIZING APPLICATION PERFORMANCE
Unifying Data and Optimizing
Application Performance with
Flash-Native, Software-Defined
Storage
Traditional storage infrastructure has become the fundamental bottleneck to performance, business process agility
and cost. Enterprises want the best of all worlds from on-premises, public cloud, and hybrid cloud solutions—across
container, VM, and bare metal workloads. Unfortunately, the required scalability, performance, and hybrid-cloud
elasticity is impossible to achieve with traditional storage arrays, hyperconverged appliances, and the first generation
of SDS (Software-Defined Storage). The ability to efficiently ingest, unify, access, and transform data underpins the
business’ ability to keep pace with its operational needs.
A Powerful Flash-Native,
Software-Defined Solution
Elastifile™ has developed a flash-native, highly
scalable, software-defined storage platform
based on the Enterprise SSD family of HGST
and SanDisk-brand products, ushering in a
new generation of software-defined solutions.
By combining the simplicity of file storage,
the scalability of object storage, and the
performance of block storage, this solution
frees administrators from the trade-offs
and limitations inherent to legacy solutions.
Whether requiring a storage solution to
accelerate HPC applications in Life Sciences
or Electronic Design Automation (EDA) to
deliver persistence for containers (e.g., via
a Docker™ platform or Kubernetes system),
or to unify data across on-premises and
cloud environments, application architects
and IT managers can now streamline and
cost-optimize their infrastructure, while
surpassing user performance expectations
with linear elasticity. With SSD capacities
increasing and $/GB cost steadily declining
over time, the transition to flash-dominated
infrastructure is well under way. The Elastifile
platform, powered by the POSIX-compliant
Elastifile Cloud File System, was designed
from the ground up, as a flash-native solution.
At the most fundamental levels, Elastifile’s
architecture was optimized to most effectively
expose the high performance of SSDs, while
also managing those devices appropriately to
maximize their endurance. However, to deliver
a truly optimized, complete solution, Elastifile
requires the use of high-performance, highcapacity, reliable storage devices. Much
as the data management infrastructure
strengthens the overall business capability,
storage hardware fortifies the capability of
the overall storage solution.
Elastifile delivers this solution based on
the Enterprise SSD portfolio of HGST and
SanDisk-brand products. The solution
delivers a combination of performance,
flexibility, scalability, and reliability to meet
the demands of the most I/O-intensive
requirements in today’s cloud enviornments.
HGST- and SanDisk-Brand
Enterprise Flash Portfolio
The HGST and SanDisk brands deliver a broad
portfolio of NVMe™-compliant PCIe, SAS, and
SATA flash drives, offering market-leading
performance, optimal price/performance
options, and capacities for software-defined
data centers and clouds.
SanDisk and HGST-brand Enterprise SSDs
enable highly scalable storage architectures
for modern enterprise deployments by
providing reliable throughput, data integrity
and protection mechanisms, and optimized
flash endurance. With capacities up to
7.68TB1 per drive, these SSDs boost results
for data-hungry applications while addressing
growing data capacity demands.
With the combined solution of the Elastifile
platform with HGST- and SanDisk-brand
SSDs, customers can choose from a range
of storage devices with the performance
and cost characteristics that best fit their
requirements, as highlighted in Table 1.
Example Reference Architecture
To demonstrate how data services can be
optimized using Elastifile, Elastifle used
SanDisk-brand SSDs to develop and test a
solution for MySQL database deployments.
The all-flash MySQL solution delivered over
70M Transactions Per Minute (TPM) and
over 1M New Orders Per Minute (NOPM)
using MySQL and standard x86 servers. This
clearly demonstrates that the combined
solution is capable of supporting scale-out,
transactional workloads to accelerate a
variety of applications. To learn more about
the MySQL solution, see the Optimizing
MySQL Performance and Scalability White
paper.
Conclusion
The combined solution of SanDisk- and
HGST-brand SSDs and the Elastifile softwaredefined storage platform provides a uniquely
powerful, all-flash, distributed storage
solution. By deploying this performanceoptimized, cost-effective platform, IT
managers can ensure that the performance
and scalability of data services do not hinder
business operations. This enables businesses
to keep pace with their operational needs and
deliver the high-quality, flexible services that
clients demand.
With the right type of flash storage from
the SanDisk and HGST brands, this solution
can deliver the performance, flexibility, and
scalability required to support the businessspecific requirements for your environment.
Visit our websites to learn more about
Elastifile, HGST-brand SSDs and SanDiskbrand SSDs.
SOLUTION BRIEF
CloudSpeed
Eco™ Gen. II
UNIFYING DATA AND OPTIMIZING APPLICATION PERFORMANCE
CloudSpeed
Ultra™ Gen. II
Ultrastar® SS200
Ultrastar® SN200
Application Workload
Read-Intensive
Mixed-Use
Read-Intensive
Mixed-Use
Read-Intensive
Mixed-use
Capacity (GB) 1
480GB – 1.92TB
400GB – 1.92TB
480GB – 7.68TB
400GB – 3.2TB
960GB – 7.68TB
800GB – 6.4TB
Form Factor
2.5” SFF
2.5” SFF
2.5” SFF
2.5” SFF
2.5” U.2
2.5” U.2
Interface
6Gb/s SATA
6Gb/s SATA
12Gb/s SAS
12Gb/s SAS
PCIe 3.0 x4 or 2x2
NVMe 1.2
PCIe 3.0 x4 or 2x2
NVMe 1.2
NAND Type
15nm MLC
15nm MLC
15nm MLC
15nm MLC
15nm MLC
15nm MLC
Sequential Read/Write (MB/s, 128KiB) 2
Up to 530/460
Up to 530/460
Up to 1,800/1,000
Up to 1,800/1,000
Up to 3,350/2,100
Up to 3,350/2,100
Random Read/Write (IOPS, 4KiB) 2
Up to 76K/14K
Up to 76K/32K
Up to 250K/37K
Up to 250K/86K
Up to 835K/75K
Up to 835K/200K
Operating Temperature
0 to 70° C
0 to 70° C
0 to 70° C
0 to 70° C
0 to 70° C
0 to 70° C
Active Power (Typ.)
3.8W
3.8W
11W
11W
25W
25W
DW/D (Random)
0.6
1.8
1
3
1
3
TBW or PBW
(Random) 5
480GB: 526 TBW
960GB: 1.05 PBW
1.92TB: 2.10 PBW
400GB: 1.31 PBW
800GB: 2.63 PBW
1.6TB: 5.26 PBW
480GB: 876 TBW
960GB: 1.75 PBW
1.92TB: 3.5 PBW
3.84TB: 7 PBW
7.68TB: 14 PBW
400GB: 2.2 PBW
800GB: 4.4 PBW
1.6TB: 8.7 PBW
3.2TB: 17.5 PBW
960GB: 1.75 PBW
1.92TB: 3.5 PBW
3.84TB: 7 PBW
7.68TB: 14 PBW
800GB: 4.4 PBW
1.6TB: 8.7 PBW
3.2TB: 17.5 PBW
6.4TB: 35 PBW
Reliability (BER)
1 in 1018
1 in 1018
1 in 1017
1 in 1017
<1 in 1017
<1 in 1017
Data Fail Recovery
√
√
√
√
√
√
Powerfail Safety
√
√
√
√
√
√
Temp. Throttling
√
√
√
√
√
√
MTBF 4
2M hrs
2M hrs
2.5M hrs
2.5M hrs
2M hrs
2M hrs
5 years
5 years
5 years
5 years
5 years
5 years
Endurance
3
Reliability
Limited Warranty 5
Table 1: SATA/SAS/PCIe SSDs
from the SanDisk and HGST brands
1.
2.
3.
4.
5.
One MB is equal to one million bytes, one GB is equal to one billion bytes and one TB equals 1,000GB (one trillion bytes) when referring to drive
capacity. Accessible capacity will vary from the stated capacity due to formatting and partitioning of the drive, the computer’s operating system, and
other factors.
Performance will vary by capacity point and/or configuration, host device, OS and application. All performance numbers are in full-sustained mode
and are peak values. 1KB=1,000 bytes (103), 1MB=1,000,000 bytes (102), 1GB=1 billion bytes (109), 1KiB=1,024bytes (210), 1MiB=1,048,576 bytes (220),
1GiB=1,073,741,824 bytes (230)
HGST-brand endurance at 100% 4KiB random-write workload. SanDisk-brand endurance at JEDEC 219 workload
MTBF (Mean Time Between Failures) based on a sample population and is estimated by statistical measurements and acceleration algorithms under
typical operating ns for this drive model. MTBF rating does not predict an individual drive’s reliability and does not constitute a warranty
The lesser of 5 years from the date of manufacture of the product or the date on which the product’s relevant endurance thresholds set forth in the
product specifications are reached. PBW=petabytes written, TBW=terabytes written
Contact Sales
North America: salesna@hgst.com | Europe/Middle East/Africa (EMEA): salesemea@hgst.com | Asia Pacific: salesap@hgst.com | Japan: salesjp@hgst.com
Western Digital Corporation
5601 Great Oaks Parkway
San Jose, CA 95119 USA
© 2017 Western Digital Corporation or its affiliates. All rights reserved. Produced 6/17.
Western Digital, the SanDisk logo, the HGST logo and Ultrastar are registered trademarks or trademarks of Western Digital Corporation
in the US and/or other countries. NVMe™ is a word mark of NVM Express, Inc. Elastifile and the Elastifile logo are registered trademarks
of Elastifile in the US and/or other countries. Other marks are the property of their respective owners. References in this publication
to SanDisk- and HGST-branded products, programs, or services do not imply that they will be made available in all countries. Product
specifications provided are sample specifications and do not constitute a warranty. Actual specifications for unique part numbers may
vary. Pictures shown may vary from actual products.
Information & Technical Support
www.hgst.com
www.hgst.com/support
Partners First Program
channelpartners@hgst.com
www.hgst.com/partners
SB37-H-ELASTIFILE-EN-US-0617-01
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising