Fujitsu ETERNUS AF650 S2, ETERNUS AF250, ETERNUS DX200 S3, ETERNUS DX500 S4, ETERNUS AF650 Configuration Manual
...
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
FUJITSU Storage
series
Non
Configuration
Non-stop storage is a high-availability solution that combines ETERNUS SF products with the ETERNUS AF
series
need to cater for future business growth and data accumulation.
ETERNUS AF series and ETERNUS DX S4/S3
-Stop Storage Reference Architecture
Guide
or ETERNUS DX S4/S3 series. It provides an ideal system environment for enterprise systems that
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
Copyright 2016-2017 FUJITSU LIMITED
Page 2 of 18
http://www.fujitsu.com/eternus/
Table of Contents
Preface 3
1. Overview of Non
2. System Configuration
3. Conclusion
-Stop Storage Solution 4
Features of Non-Stop Storage Solution 4
1.1.
Suitable Systems for Non-Stop Storage 4
1.2.
Protecting Investment and Optimizing Investment Cost 5
1.3.
FUJITSU Storage ETERNUS SF Software for Non-Stop Storage Operation 5
1.4.
7
Example of a Non-Stop Storage Configuration 8
2.1.
Example of a Non-Stop Storage Configuration (with Redundant Management Servers) 12
2.2.
Hardware Layout 13
2.3.
Workflow for System Configuration 15
2.4.
18
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
Copyright 2016-2017 FUJITSU LIMITED
Page 3 of 18
http://www.fujitsu.com/eternus/
Preface
Non-stop storage is a high-availability solution for business continuity.
A non-stop operation is possible during problems and disasters with a cluster configuration created by combining FUJITSU Storage ETERNUS
SF products with the FUJITSU Storage ETERNUS AF series or ETERNUS DX S4/S3 series. This document provides an overview of the non-stop
storage solution and describes its configuration.
Prerequisites
The product lineup and product information stated in this document are current as of December 2017.
Target Readers
This document is intended for readers familiar with FUJITSU Storage ETERNUS SF products, the FUJITSU Storage ETERNUS AF series and
ETERNUS DX S4/S3 series, VMware vSphere, and Windows Server.
Target Models
This document relates to the following FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series models.
- FUJITSU Storage ETERNUS AF250 S2/AF650 S2 and ETERNUS AF250/AF650
- FUJITSU Storage ETERNUS DX200 S4/S3
- FUJITSU Storage ETERNUS DX500 S4/DX600 S4 and ETERNUS DX500 S3/DX600 S3
- FUJITSU Storage ETERNUS DX8100 S3
- FUJITSU Storage ETERNUS DX8700 S3/DX8900 S3
Abbreviations
This document uses the following abbreviations.
- FUJITSU Storage ETERNUS AF series … ETERNUS AF series
- FUJITSU Storage ETERNUS DX S4/S3 series … ETERNUS DX S4/S3 series
- FUJITSU Storage ETERNUS AF series and
ETERNUS DX S4/S3 series … ETERNUS AF/DX
- FUJITSU Storage ETERNUS SF Storage Cruiser … ETERNUS SF Storage Cruiser
- FUJITSU Storage ETERNUS SF products … ESF
- ETERNUS SF Manager … ESF Manager
- ETERNUS SF MA … MA
- Automated Storage Tiering function … AST
- Automated Quality of Service … Auto QoS
- VMware vSphere … vSphere
- VMware ESXi … ESXi
- VMware vCenter Server … vCenter Server
- Microsoft Windows Server 2012 R2 … Windows Server
- Windows Server Failover Clustering … WSFC
ctive Directory Server … AD Server
- A
- Domain Name System Server … DNS Server
- Network Time Protocol Server … NTP Server
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
Copyright 2016-2017 FUJITSU LIMITED
Page 4 of 18
http://www.fujitsu.com/eternus/
Fluctuating operation
Growing business
No business downtime
Continuous operation,
life cycle
ETERNUS AF/DX
1. Overview of Non-Stop Storage Solution
Redundancy has long been adopted for servers, with this also providing redundancy for any internal hard disks in the server. While
configurations such as the shared disks in cluster systems or server virtualization have involved servers using external disk storage,
redundancy for the storage systems themselves has relied on the server software and its use has remained limited.
In recent years, however, awareness of business continuity has been rising and virtualization involves running many virtual machines on the
same physical storage system. For these reasons, redundancy is also required for general-purpose storage systems.
Non-stop storage means a storage environment that can maintain uninterrupted operation with high reliability and optimal performance
over long periods of time, even in the event of physical failures, problems, or emergencies. Locating physical servers and storage systems in
different buildings or at different sites provides an ideal environment for mission-critical systems because they can remain in operation
during emergencies such as problems or power outages.
Features of Non-Stop Storage Solution 1.1.
This solution provides redundancy for storage systems to provide an environment in which the customer’s business can continue to operate
even if physical damage occurs as a result of a problem or emergency by automatically switching the storage systems. Moreover, data
consistency and the ability to recover are maintained even if data corruption is present after such a switchover because the backup data is
automatically carried over, along with business data.
This solution includes Automated Storage Tiering (AST) that has the ability to automatically select a suitable storage device for the storage
cost and the Automated Quality of Service (Auto QoS) function that automatically optimizes access to the storage area network (SAN) for
linking servers with storage systems. This provides a reliable storage environment that can both maintain performance and reduce storage
costs despite fluctuations in workload, business expansion, or the accumulation of data over time.
Suitable Systems for Non-Stop Storage 1.2.
This solution is suitable for the following systems.
- Systems that require high reliability with uninterrupted 24-hour/365-day operation
Systems that must remain in operation even if storage systems are unavailable due to equipment failure or building power failure, etc.
- Systems that demand stable storage performance
Systems that are subject to variable loads of access to storage systems and virtualization environments that host various business
systems
- Systems where workloads are expected to increase over time
Systems on which storage capacity requirements and throughput continue to increase after installation due to business growth
The following shows application examples of this solution.
- Application to financial systems
The solution provides non-stop operation with data redundancy across storage systems for systems that provide 24-hour/365-day
services, such as Internet banking or ATMs, including during electrical maintenance inspections, storage failures, and disasters.
- Application to systems for managing medical imaging data
The solution provides data redundancy at the storage system level for medical imaging data, the safe-keeping of which is a legal
obligation, and reduces storage costs by automatically moving data for recovered patients to low-cost Nearline disks.
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
Copyright 2016-2017 FUJITSU LIMITED
Page 5 of 18
http://www.fujitsu.com/eternus/
Protecting Investment and Optimizing Investment Cost 1.3.
By using storage virtualization to start small and expand as required, this solution provides a storage environment that protects the
customer’s initial investment and can flexibly deal with business growth even for systems where future expansions are difficult to estimate.
Storage scalability typically means capacity expansion and throughput improvement.
Capacity expansion is achieved by adding drives. If more expansion is needed, additional drive enclosures can be added to provide space for
more drives. If a large amount of capacity is required for reasons such as business growth, additional storage systems can be installed.
These three ways of adding capacity provide seamless capacity scalability in response to business growth.
In the case of IOPS performance improvement, the IOPS for a specific operation can be upgraded by using flash memory to provide a larger
cache or by selecting high-speed drives (SSDs).
The cache size can be increased up to 800GB for the entry-level ETERNUS DX200 S4 storage system that is used in the example for this
solution or up to 5.6TB for an enterprise storage system.
Furthermore, the addition of a small-capacity SSD can increase IOPS by using AST. This function automatically reallocates frequently accessed
areas to SSDs to provide performance in the way that best suits the operation.
To improve the throughput per unit time, additional ETERNUS AF/DX storage systems can be installed without business downtime. Installing
additional storage systems dramatically enhances the throughput.
In this way, the solution protects the customers’ investment by making the most of their initial investment and by allowing additional
investment to be made in small increments.
FUJITSU Storage ETERNUS SF Software for Non-Stop Storage Operation 1.4.
The FUJITSU Storage ETERNUS SF storage software is used by this solution to provide non-stop operation and to handle the operation and
management life cycle. With an extensive product range that extends from resource management of storage systems to the provision of
additional functions such as replication, it flexibly supports administration of the storage environment and system-wide optimization.
The following sections describe the features of the ETERNUS SF products that provide powerful support for the reliable operation of the
storage systems centered on the ETERNUS AF/DX.
ETERNUS SF Storage Cruiser 1.4.1.
ETERNUS SF Storage Cruiser
system.
Used together with options such as AST *2 and Auto QoS
*1
provides integrated management of the storage environment to support reliable operation of the entire
*3
, it provides cost reductions and more efficient use of resources.
The main functions are as follows.
- Simplified storage design and configuration
A simple GUI is provided for easy configuration, from volume and virtual storage pool creation to allocation to servers.
Users can start small with the minimum physical drive capacity and later add drives without interrupting operation.
- Visualization of storage resources
A dashboard displays information of interest, enabling the status of the storage system and any problems to be identified early.
Performance graphs provide the information for optimizing the I/O performance needed to satisfy business requirements.
It provides integrated management of the components, including the storage system, Fibre Channel switches, physical servers, and
virtual servers using Hyper-V, VMware, or other software.
- Use of AST to optimize investment
AST is used in environments that combine drives of different types to automatically reallocate data to the most appropriate drive based
on access frequency.
It optimizes storage costs by allocating frequently accessed data to SSDs or other ultra-high-speed drives to maintain performance, and
allocating less frequently accessed data to inexpensive Nearline disks with high capacity.
- Use of Auto QoS to allocate performance resources
For systems where the storage system is being shared by a number of operations, Auto QoS prioritizes the I/Os of priority operations
thereby ensuring that high priority operations operate reliably without being affected by fluctuations caused by I/O loads of other
operations.
This makes efficient use of resources by allowing business data with a wide range of different priorities to be combined.
Response times can also be reduced by using automatic tuning together with AST.
The ETERNUS AF/DX and ETERNUS SF Storage Cruiser together provide an ideal system for maintaining business continuity that can be
used safely and reliably.
FUJITSU Storage ETERNUS AF series and ETERNUS DX S4/S3 series
Copyright 2016-2017 FUJITSU LIMITED
Page 6 of 18
http://www.fujitsu.com/eternus/
Storage Cluster 1.4.2.
Storage Cluster
*4
is an optional transparent failover feature of ETERNUS SF Storage Cruiser, which enables a non-stop storage infrastructure in
the event of a single point of failure (SPOF) in a storage system.
To improve system availability, it automatically detects when the storage system recovers so as to restore redundancy as soon as possible.
The main functions are as follows.
- Data mirroring provided by storage functions
The synchronous Advanced Copy function of the ETERNUS AF/DX maintains data consistency across different storage systems.
Copying does not use server resources, thereby keeping any load on the operation to a minimum.
- Automated monitoring, switchover, and post-recovery switch-back
When a failure is detected in the ETERNUS AF/DX, Storage Cluster switches the access destination automatically without operator
intervention.
Even when AST is running, data allocation is still optimized on the Secondary storage based on the access information on the Primary
storage. This means that the same I/O performance is maintained after a switchover occurs and the operation can continue to function
reliably.
Post-recovery procedures are simplified by automatically detecting a data consistency restoration after the storage system recovers and
the access destination switches back.
- Continuity of replication
When replication is used on the Primary storage, the same copying is also performed on the Secondary storage at the same time.
Because copying continues to function after switchover, replication continues and its high reliability is maintained.
Using Storage Cluster provides a system without business downtime.
Because automatic switchover does not require any operation, it helps minimize the operating costs of the customer’s system.
ETERNUS SF MA 1.4.3.
ETERNUS SF MA collects and archives performance data, and can output information such as automatic diagnosis results or long-term
performance, not only on screen but also in the form of customizable reports.
The main functions are as follows.
- Automatic diagnosis
Problems with performance can be pre-empted through automatic diagnosis performed using indicators determined based on expert
know-how and (customizable) threshold values.
Diagnosis results can be scheduled for automatic output in regular reports.
Diagnosis results can also be presented as Baseline reports output in Excel format.
- Extensive performance data display and Navigation pane
The relationship between the resources can be analyzed from the server side or the storage system side by tracking the resources in a tree
structure.
This provides intuitive operation of display and performance analysis, enabling bottlenecks or the scope of problems to be identified
quickly.
For example, ETERNUS SF MA interoperates with Microsoft Windows Server, VMware vCenter Server, and Hyper-V to undertake
investigations that link the storage system and the IP network switches on the paths with performance problems for business servers by
displaying the storage access performance from the physical servers and the virtual machines.
- Storage cost reduction by using compression and high-speed queries
ETERNUS SF MA uses Fujitsu's highly efficient data compression for long-term storage of performance and event information.
Performance analyses covering long periods of time can be performed quickly using in-memory processing.
The solution provides a storage system that can operate non-stop, 24 hours a day, 365 days a year.
The performance data display and automatic diagnosis functions of ETERNUS SF MA are useful to identify long-term performance trends and
prevent performance problems arising due to business growth.
Use of the Navigation pane and reports of ETERNUS SF MA also reduces administrator workloads.
*1: Requires the ETERNUS SF Storage Cruiser license for each ETERNUS AF/DX.
*2: Requires the ETERNUS SF Storage Cruiser Optimization option for each ETERNUS AF/DX.
*3: Requires the ETERNUS SF Storage Cruiser QoS Management option for each ETERNUS AF/DX.
*4: Requires the ETERNUS SF Storage Cruiser Storage Cluster option and the ETERNUS SF AdvancedCopy Manager license for each ETERNUS AF/DX.
Loading...