HPE Apollo 4000 Systems
with Scality RING Scalable Storage
Overview
HPE Apollo 4000 Systems with Scality RING Scalable Storage
HPE Apollo 4000 Systems with Scality RING Scalable Storage is a scale-out software-defined storage platform that runs on HPE
effective data lakes hosting
) protocols, easily integrating into many standard
The software enables continuous availability through hardware failures, capacity expansions, software
both hardware and
software entitlements in one easy-to-order product.
Support for RING8 XDM and Scality Zenko, for hybrid cloud data management:
• Enabled stand-alone with Zenko (RING not required)
Apollo 4000 Servers and is designed for multi-application environments needing to store unstructured data, at petabyte scale.
HPE and Scality RING, Enterprises and Service Providers building cloud storage solutions can create costa nearly unlimited number of items in a single name-space, instead of the data silos creates by legacy storage appliances.
Scality RING Scalable Storage employs both replication and erasure coding, across single or multiple sites,
achieve up to 14 nines (9s) of data durability with higher storage efficiency, requiring less raw storage and therefore less cost
comparable solutions. Scality RING supports object (S3/REST) and file (NFS/SMB
and custom applications.
upgrades, and hardware upgrades, enabling customers to take advantage of future innovations in the HPE Apollo 4000
including performance, capacity, components, and architecture.
HPE offers this solution on the HPE Apollo 4000 storage systems which are specifically optimized for storage-centric workloads
Customers may purchase the software separately from the hardware or may select a bundled product combining
• A single unified console to manage data across multiple clouds offering
- Searchable metadata across multiple clouds
- Policy-driven data lifecycle management for 1-to-many copies, data transition (movement), and space-freeing
data deletion
• Enabled with RING systems with RING8 XDM (requires RING 7.4.4 or greater)
Page
HPE Apollo 4000 Systems with Scality RING Scalable Storage
Standard Features
Key Features of Scality RING
Optimal performance on HPE Apollo 4000 storage servers
Scality have certified RING on the 24LFF HPE Apollo
4200 Gen10 and the 60LFF Apollo 4510 Gen10 servers. Apollo 4000 systems are purpose built for big data analytics, software
to help customers meet their data center operation
is simpler and deployments are highly adaptable. Overall, this
significantly reduces acquisition and operating costs and increases project flexibility.
Supports mixed applications & workloads
s can
simply leverage other interfaces (such as object) in the future.
Continuous uptime through hardware failures and maintenance
ns,
with minimal intervention. This enables carrier-grade service levels and improved end user satisfaction.
Scales to petabytes and beyond
additional disparate components. This enables massive
consolidation and significantly reduces operating costs.
pool, and data
is concurrently accessible by file and object interfaces.
Peer-to-peer Shared Nothing Design
Unique RING topology with peer-to-peer architecture & scalable routing mechanism.
Designed to keep running through hardware failures, hardware refreshes, capacity upgrades, and software upgrades.
The RING environment may be deployed to a single site, stretched over multiple sites, or replicated locally and/or across sites.
Advanced & Flexible Data Protection
based replication, for efficient storage of
small objects. Both erasure coding and replications (up to 6 copies) may be used simultaneously in the same RING cluster.
Scality RING software is designed to be used with industry standard x86 servers with support for SSD & HDDs. To optimize you
business performance with object storage, Hewlett Packard Enterprise and
defined storage, backup and archive, and other data storage-intensive workloads.
challenges
By deploying your Scality RING solution on HPE qualified platforms, customers can rapidly adopt new hardware and media innova
with no need for data migration on hardware refresh. Support
Scality has the performance and latency characteristics to manage mixed storage workloads in a single pool that encompasses f
and S3-compatible object applications. This enables data consolidation to increase utilization and economies of scale and elimi
storage silos.
Customers can reduce time to market by integrating existing applications using standard protocols and rich API's. Application
-
Customers can maintain availability through hardware failures, capacity expansions, software upgrades, and hardware generatio
A RING storage infrastructure scales out linearly as a single system across multiple active sites, thousands of servers, hund
petabytes, and unlimited objects without adding administrators or
No storage silos: object storage technology abstracts underlying storage servers to create a uniformly scalable storage
Both the access layer and the storage layer are independently scalable to thousands of nodes, all of which can be accessed di
and concurrently.
Automatically utilizes both Erasure Coding, to reduce overhead for large objects, and policy-
Page
HPE Apollo 4000 Systems with Scality RING Scalable Storage
level, and asynchronous
mirrored clusters for higher-latency DR configurations.
Customizable Availability and Failure Domains
site configurations) and multiple
rack/server/disk failures.
Self-healing
Self-healing after component failures.
Rich Management Portal and CLI
Simple Operations & Management
line quickly, supported with HPE Apollo servers and Scality
Distributed Object Storage
Scality S3 Connector, compatible with AWS S3 and IAM security services. May be integrated with Microsoft Active Directory
o theoretical size limit on objects, however AWS S3 API dictates Multipart upload objects are are limited to 10000 parts of
maximum 5GB each, for a total of 50TB per object
key values, file system metadata and POSIX
• Space quotas for NFS & SOFS
Active/Active storage pools across physical data centres, Active/Active selective replication at the S3 bucket-
User-defined policies to achieve 14 nines (9's) of data durability, tolerance for whole-site (in multi-
Full management and control with Supervisor and CLI
• Full Topology View: RING status with Zone, Server & Node details
SNMP MiB support and graphical usage monitoring
• System capacity and performance expansion
• Automated disk failure detection and handling framework
• Automatically rebuild failed drive data on remaining drives
• Automated storage rebalancing
• Automated failed disk replacement to bring new hardware on-
RING versions 7.4 and greater
• Scality HTTP/REST (sproxyd)
•
or other enterprise directory software that supports the SAML protocol. Sproxyd native REST API
• Virtually unlimited object capacity
• N
• Fully parallel and POSIX compliant file system
• FUSE (sfused) connector, local Linux host file system access to RING storage services
• NFSv3 connector
• NFS Ganesha connector (RING8) supporting NFSv3 and NFSv4, IPv6, Kerberos Authentication, and POSIC ACL
• SMB 2.0/3.0 connector
• CDMI REST connector, SNIA CDMI 1.0.1 compliant
• IPV6 support for NFS and SMB
• Persistent shared cache across physical connectors
• Software load-balancing supporting continuous ingest during connector failure
• Unique internal distributed database (MESA) – for full scale-out support of object
methods, and block volumes
• Kerberos KDC/Microsoft Active Directory server-based authentication support for NFSv3 and SMB 2.0/3.0
HPE Apollo 4000 Systems with Scality RING Scalable Storage
provide the interface to business applications. Connectors and RING storage are independently
scalable. Connectors may be run on the RING storage systems, on VMs, or on separate servers such as ProLiant DL360
• eXtended Data Management as a modular feature, to be deployed outside of RING
RING 8 Extended Data Management (XDM) and Scality Zenko
Performance on Apollo 4000
o Connector access points can scale independently of storage by using external connector servers
hour, support for the Scality RING software solution. Customers are provided with complete installation
access to advanced services such as proactive health checks and migration assistance. With purchase of HPE
, HPE Apollo 4000 systems, and HPE Pointnext operational support such as Foundation Care Service or
call response support handling when contacting HPE for product support, with HPE
support tracking through to problem resolution.
RING file and object connectors
• Distributed Lock Manager (DLM) for multiple SOFS connectors over the same SOFS volume
• New NFS Ganesha connector supporting NFSv4 and IPV6, Kerberos authentication
• Non-root execution for all RING components
• S3 connector with KMIP for KMS and S3 Quotas
• Data orchestration and mobility across RINGs with no additional licensing required
• May be extended to other clouds with purchase of additional data licenses
• Lifecycle policies, 1-to-many replication, transition, and expiration
• Searchable global namespace
• RING8 XMD sold with data licenses (one-time purchase), Scality Zenko sold with data subscriptions (renewable annually)
• Object performance:
− Writes: up to 850 MB/sec on very large files (50 MB and greater)
− Reads: up to 1.3 GB/sec on very large files
− Additional performance characterization available upon request
• File performance:
− With NFS connector, up to 900 MB/sec reads and writes on very large files
− With SMB, up to 400MB/sec reads and writes on very large files
• Other performance
Near-linear scalability across multiple connector types
−
Scality provides worldwide, 24services and given
Scality RING software
Datacenter Care Service, customers get first-
HPE Apollo 4000 Systems with Scality RING Scalable Storage
HPE Apollo 4000 storage servers and fast start bundles are validated designs supported by the Scality Fast Track program to
e deployed on ProLiant DL360
based Cloudline CL5800. For more information configuring Scality RING software
age servers should be tailored to the specific workload requirement. Data protection strategy and average
sentative to document your individual solution needs and design an environment with required services
to deploy.
HPE Apollo 4510 Gen10 Systems
Apollo 4510 systems
supports up to 60 LFF drives for bulk storage, with solid state options in both LFF carriers and as NVMe cards in PCIe slots.
HPE Apollo 4200 Gen10 Server and fast start bundle
optimized for storage density, performance, configuration flexibility and
affordability. Solid state options are available in both LFF carriers and as NVMe cards in PCIe slots.
speed design processes through standardized configurations. A Scality RING may be deployed with as few as three storage serv
For customers who have standardized on other HPE platforms, Scality RING software from HPE may b
with D8000 disk enclosures, or on the open designwith HPE ProLiant or HPE Cloudline, contact your HPE sales representative.
NOTE: Sizing for stor
file size should be taken into account for optimal configuration. As such it is recommended that you engage with your Hewlett
Packard Enterprise repre
The HPE Apollo 4500 systems are purpose-built for maximizing bulk storage in 4U of rack space. The HPE
The HPE Apollo 4200 Server is a 2S/2U storage server