The IBM TotalStorage SAN File System -- What It Does
The SAN File System is designed as a highly available file system for
SAN-attached storage that provides file sharing and centralized storage
management for UNIX, Windows, and Linux servers. The SAN File System is
intended to combine the benefits of file sharing across servers provided
by Network Attached Storage (NAS) with the benefits from high performance
data access provided by Storage Area Networks (SAN).
The SAN File System is designed to support high performance data
sharing for heterogeneous servers accessing SAN-attached storage. It has
a common file system for UNIX, Windows, and Linux servers, with a single
global namespace providing data sharing across servers. The SAN File
System is designed to support the local file system interfaces on UNIX,
Windows, and Linux servers, and to be used typically without requiring
any changes to your applications or databases that use a file system to
store data. The SAN File System is designed as a highly scalable
solution supporting both very large files and very large numbers of files
without the limitations normally associated with Network File System
(NFS) or Common Internet File System (CIFS) implementations.
The SAN File System is designed to help lower the cost of storage
management and enhance productivity by providing centralized storage and
data management for supported servers using the SAN File System. It
provides policy-based storage management through the use of customer-
defined rules that specify what storage is used when a file is created.
With SAN File System V2.2, you can define rules that specify when a file
should be moved between different storage pools and also can specify when
a file should be deleted. Storage can be virtualized into storage pools
organized by service class, and files can be automatically placed on the
appropriate storage resource, thereby, helping reduce administrative
The SAN File System has the potential to reduce the overall amount of
storage required in many environments by allowing storage resources to be
shared more efficiently across servers. A common pool of free space and
temporary space can be shared across all of the servers, helping to lower
the amount of free space and temporary space that is required. In
addition, because files can be shared across servers, storage can be
saved by reducing the need to maintain duplicate copies of files.
The SAN File System also has the potential to improve application
availability by reducing or eliminating application downtime for storage
and data management tasks. It allows storage volumes to be added,
replaced, or removed without disrupting application processing.
Application downtime for data backups can be reduced through the use of
the file-based FlashCopy function. Application servers can be added on
demand to the SAN File System allowing non-disruptive growth in
SAN File System Software
SAN File System V2.2 will be delivered as a software-only program
product. The hardware requirements for the SAN File System metadata
servers and for the master console are defined in the Hardware
Requirements section. The SAN File System metadata servers run as
a cluster designed for high availability. The cluster consists of a
minimum of two, and up to a maximum of eight servers running SAN File
System metadata servers.
The SAN File System software consists of three components: the
software that runs on the SAN File System metadata servers, the software
that runs on your application servers, called the SAN File System Client
software, and the software the runs on the Master Console.
The SAN File System metadata servers provide metadata, information
about the files, to the SAN File System Client software running on your
application servers. Once the SAN File System Client software has the
metadata, then the SAN File System Client software accesses the file data
directly via the SAN.
The SAN File System metadata servers also provide the management
support for the SAN File System. The SAN File System software includes a
task-oriented GUI that can be accessed from a Web browser, a command line
interface that can be used to create scripts for frequently used
commands, and an application programming interface (API) that can be used
by other storage and data management applications to manage or monitor
the SAN File System. The API is based on the Common Information Model
(CIM) and is encoded in XML with common methods and syntax to the Storage
Networking Industry Association's (SNIA) Storage Management Interface
The SAN File System Client can be run on application servers. SAN
File System V2.2 supports application servers running Microsoft Windows
2000 Server, Microsoft Windows 2000 Advanced Server, Microsoft Windows
Server 2003 Standard Edition, Microsoft Windows Server 2003 Enterprise
Edition, IBM AIX V5.1 (32-bit), IBM AIX V5.2 (32-bit and 64-bit), AIX 5.3
(32-bit and 64-bit), Red Hat Enterprise Linux Server 3.0. (32-bit), SuSE
Linux Enterprise Server 8.0 (32-bit), and Sun Solaris 9 (64-bit).
The SAN File System supports application servers running in a
clustered environment. SAN File System V2.2.0 supports application
servers with AIX running with HACMP and Sun Solaris servers running in a
clustered environment. SAN File System V2.2.1 supports application
servers with Microsoft Windows 2000 and Microsoft Windows 2003 running in
a clustered environment.
The SAN File System is designed to support the local file system
interfaces on UNIX, Linux, and Windows servers. This means that the SAN
File System is designed to be used without requiring any changes to your
applications or databases that use a file system to store data. The SAN
File System Client for AIX, Solaris, and Linux use the virtual file
system interface, within the local operating system, to provide file
system interfaces to the applications running on IBM AIX, Sun Solaris,
Red Hat Linux and SuSE Linux. The SAN File System Client for Microsoft
Windows uses the installable file system interface within the local
operating system to provide file system interfaces to the applications
running on Microsoft Windows 2000 Server, Microsoft Windows 2000 Advanced
Server, Microsoft Windows Server 2003 Standard Edition, and Microsoft
Windows Server 2003 Enterprise Edition.
The master console software is designed to provide a unified point of
service for the entire SAN File System cluster, simplifying access to,
and administration of, metadata servers. The master console has Remote
Support and IBM Service Alert (Call Home) capabilities to provide fast,
efficient, secure service. It makes a Virtual Private Network (VPN)
connection readily available that you can initiate and monitor to enable
hands-on access by remote IBM support personnel. It also provides a
common point of residence for the IBM Tivoli Storage Area Network
Manager, IBM Director, and other tools associated with the capabilities
just described, and can act as a central repository for diagnostic data.
The SAN File System provides the following functions:
- Uniform global namespace
- Storage pools
- Policy-based file space provisioning
- Storage quota management
- Volume drain
- FlashCopy image
- File copy migration tool
Uniform global namespace: All supported application servers
connected to the SAN File System can access the same data using SAN File
System's uniform global namespace. A uniform global namespace provides
the capabilities for all clients to have a consistent view of the SAN
File System directory and files which can enable enhanced workflow and
collaboration between applications running on multiple servers.
Storage pools: The SAN File System provides the capability
to assign storage volumes to storage pools where the storage pools can
represent different classes of service for storage. Volumes can be added
to a storage pool without impacting application processing. Data can be
moved between volumes without impacting availability or performance.
Policy-based file space provisioning: Storage
administrators will be able to create rules for the SAN File System to
determine what storage pool is used to allocate space for a file when the
file is created. The storage administrator can use any of the file
attributes (such as file name, file type, date created, user ID or group
ID, and so on) to create these rules. Through the use of the policy
based rules, the SAN File System automates the task of allocating space
on the desired storage volumes.
Storage quota management: The SAN File System provides
storage capacity to applications and servers creating new files or
extending existing files. The SAN File System allows the storage
administrator to define the amount of storage to be used by a specific
server or by a specific application (defined as a fileset) and can help
the storage administrator monitor the storage utilization by alerting
them when the storage reaches a pre-defined capacity. The use of storage
quotas can prevent an application from using more storage than permitted.
The SAN File System allows for storage utilization quotas to be set
higher than the physical storage available to the system or applications.
This allows for simplified management of the system by allowing
administrators to dynamically add storage without having to reset
application level policies. Both hard and soft quotas are supported
along with threshold alerts for integration with systems management
Locking: The SAN File System is designed to provide file
level locking to the SAN File System clients running on your application
servers. This function supports data sharing with integrity across
heterogeneous application servers. The file level locking uses the same
locking permissions (read, write, execute, and so on) that are provided
by the local operating systems for locking within one server. In
addition to locking the entire file, the SAN File System also supports
the ability to do a lock on a byte range within a file.
Caching: The SAN File System clients cache the metadata
that they get from the SAN File System metadata servers. This allows the
applications running on a server to continue to access the file data
without additional requests to the metadata servers and also allows other
applications on the same server to share files without additional
requests to the metadata servers. The caching of the metadata supports
high performance data access over the SAN without requests flowing to the
metadata server. It also supports continuous access to open files even
if a temporary condition prohibits communication with the metadata
Volume drain: The SAN File System provides a volume drain
function that allows you to remove or upgrade storage volumes without
disrupting application processing. The volume drain function moves the
data off of the specified volume onto other volumes in the same storage
pool. The SAN File System manages the access to the data that is being
moved so there is no need to stop application processing during the
volume drain. Data can be moved between volumes typically without
impacting availability or performance.
Flashcopy image: The SAN File System provides a file-based
point-in-time copy function called FlashCopy image. The FlashCopy image
is a logical point-in-time image of a Fileset, a set of files. A copy of
the data block is made upon the next update after the FlashCopy image.
The FlashCopy image only requires storage for changed blocks, therefore,
reducing the amount of storage needed. The SAN File System supports 32
FlashCopy images for each Fileset. The files can be restored to any one
of the FlashCopy images.
File-based backup and restore products can back up files using the
FlashCopy image while the applications are online and updating a later
version of the files. The file-based FlashCopy image can help
significantly reduce maintenance windows for data backups and can also
help reduce the amount of storage space needed for on-line data backups.
File copy migration tool: The SAN File System provides a
file copy migration tool that you can use to copy files into the SAN File
System. The file copy migration tool runs on your application servers
that are the SAN File System clients. The file copy migration tool
includes a planning option that creates an estimate of the time needed to
copy the files into the SAN File System.
New Functions in SAN File System V2.2.0
- Support for policy-based movement of files between storage pools and
policy-based deletion of files.
- Support for file sharing with cross platform user authentication and
security permissions between Windows and UNIX environments.
- Support for the SAN File System clients on SuSE Linux servers.
- Support for iSCSI attached clients and storage.
- Support for additional globalization support including Unicode
fileset attach point names and Unicode file name patterns in policy
Policy-Based Movement of Files Between Storage Pools
The SAN File System provides basic file lifecycle management allowing
administrators to specify how files should be automatically moved among
storage pools during their lifetime, and optionally, specify when files
should be deleted. This feature can potentially lower the overall costs
of storage and improve storage space utilization allowing a balanced use
of premium and inexpensive storage matching the objectives of the
enterprise. This feature also has the potential to reduce manual storage
administration for managing space utilization and can reduce the cost of
The Move File feature enables administrators to non-disruptively
transfer the contents of an individual file, along with any FlashCopy
images of the file, from a source pool to a destination pool. The file
is de-fragmented and distributed across available volumes in the
destination pool as its blocks are transferred. All FlashCopy images of
the given file are moved to the target pool along with the file itself.
Applications do not need to be quiesced during the move.
Heterogeneous File Sharing Between Windows and UNIX Servers
The Heterogeneous File Sharing feature enables flexible, yet secure,
sharing of files between UNIX and Windows servers. The SAN File System
provides cross-environment access checking such that files created on
Windows can be accessed by authorized users on UNIX, and vice versa. To
control cross-environment authorization, the administrator manages a set
of user map entries using the administrative CLI or GUI. Each user map
entry specifies a UNIX domain-qualified user and a Windows domain
qualified user that are to be treated as equivalent for the purpose of
checking file access permission in the cross-platform case. The SAN File
System metadata server cluster accesses the customer's UNIX and Windows
directory services, as needed, to obtain user ID and group membership
iSCSI Attached Clients and Storage
The SAN File System supports host application servers and storage
devices attached through iSCSI. The iSCSI attached storage devices must
conform to the SCSI driver interface standards, including the unique
device identification calls.
New Function in SAN File System V2.2.1
- Supports the SAN File System Client software on Microsoft Windows
2000 and Microsoft Windows 2003 running in a clustered environment.
Section 508 of the U.S. Rehabilitation Act
SAN File System software is capable, when used in accordance with
IBM's associated documentation, of satisfying the applicable requirements
of Section 508 of the Rehabilitation Act provided that any assistive
technology used with the product properly interoperates with it.
Back to top
SAN File System software requires hardware for the SAN File System
metadata server cluster and hardware for the master console. The SAN
File System supports two to eight SAN File System metadata servers
running in a cluster.
SAN File System V2.2 metadata server software is supported on the SAN
File System Metadata Server engine that was previously available with SAN
File System V1.1 (4146 Model 1RX) after the operating system has been
upgraded to SuSE Linux Enterprise Server 8.0. The SAN File System V2.2
metadata server software is also supported on IBM eServer xSeries servers
or equivalent servers meeting the requirements defined below. SAN File
System V2.2 will support a cluster of metadata servers consisting of both
4146-1RX engines and IBM eServer xSeries servers or equivalent servers.
If not using the SAN File System Metadata Server engine, the following
hardware components are required for each metadata server:
- A high performance, highly reliable Intel processor server: IBM
eServer xSeries 345 or 346 server or equivalent Intel server with the
- Two processors of at least 3 GHz.
- At least 4 GB system memory.
- Dual power supplies are recommended but not required.
- Dual internal hard drives with mirroring (at least 36 GB).
- CD-ROM drive.
- Diskette drive.
- One 10/100/1000 megabit port for Ethernet connectivity (Copper or
- Two Fibre Channel Host Bus Adapter (HBA) ports supporting 2 Gb speed.
- Remote Supervisor Adapter 2 (RSA II) card.
- Certified for SUSE Linux Enterprise Server 8, kernel level
SAN File System V2.2 supports the Master Console hardware that was
previously available with SAN File System V1.1 (4146-T30 feature #4001).
In addition, SAN File System V2.2 Master Console software will be
supported with specific IBM eServer xSeries servers or equivalent servers
as defined below.
If not using the SAN File System hardware (4146-T30 feature #4001),
the following hardware components are required for the Master Console:
- A high performance, highly reliable rack-mounted Intel Pentium 4
processor server: IBM eServer xSeries 305 or 306 server or equivalent
Intel server with the following capabilities:
- At least 2.6 GHz processor speed
- At least 1 GB system memory
- Two 40 GB IDE hard disk drives
- CD-ROM drive
- Diskette drive
- Two 10/100/1000 megabit ports for Ethernet connectivity (Copper or
- Two Fibre Channel Host Bus Adapter (HBA) ports
- Monitor and keyboard: IBM Netbay 1U Flat Panel Monitor Console Kit
with keyboard or equivalent
- KVM Switch: Keyboard, Video, Mouse Switch or equivalent
Note: The master console must be in proximity with the SAN File System
server. If such location is not feasible, the customer must supply an
extra keyboard and display for SAN File System access to the master
For additional information on part numbers for the hardware
prerequisites, refer to the SAN File System Support Web site, at:
Supported Storage Devices
SAN File System V2.2 can be configured with any SAN storage device for
the user data storage. That is, any SAN-attached storage supported by
the operating systems on which the SAN File System client runs can be
used provided it conforms to the SCSI standard for unique device
SAN File System V2.2 can be configured with the following storage for
the metadata storage:
- IBM TotalStorage Enterprise Storage Server (ESS) Models F20, 750, and
- IBM TotalStorage SAN Volume Controller V1.1 or V1.2 with the storage
devices supported by the SAN Volume Controller
- IBM TotalStorage DS4300 Turbo, DS4400, and DS4500 (formerly called
IBM TotalStorage FAStT Models 600T, 700, and 900)
For updated and current support levels, refer to the SAN File System
Support Web site, at:
Requests for additional types of storage devices for metadata storage
will be accepted through the TotalStorage Open Software Family "One Off"
process. Contact your TotalStorage Open Software Sales representative to
submit a "One Off" nomination form.
The SAN File System V2.2 requires the following software to be
installed on the metadata servers:
- SUSE Linux Enterprise Server 8, kernel level 2.4.21-231, (SP3).
- Multi-pathing driver for the storage device used for the metadata
LUNs. Either SDD 1.5.1 or RDAC 8.41 are required depending on the
storage device used for the metadata LUNs.
The SAN File System requires the following software to be installed on
the Master Console:
- Microsoft Windows 2000 Server Edition (SP3, or later) or Microsoft
Windows Server 2003 Standard and Enterprise Editions.
- Microsoft Windows Internet Explorer Version 6.0 (SP1, or later).
- Anti-virus software is recommended.
SAN File System supports connectivity with Intel-based servers running
Microsoft Windows 2000 Server, Microsoft Windows 2000 Advanced Server,
Microsoft Windows Server 2003 Standard Edition, Microsoft Windows Server
2003 Enterprise Edition, Red Hat Linux, or SuSE Linux and RISC-based
systems running IBM AIX or Sun Solaris. The SAN File System V2.2 clients
run on a supported version of the following operating systems:
- Microsoft Windows 2000 Server with Service Pack 4, or later.
- Microsoft Windows 2000 Advanced Server with Service Pack 4, or later.
- Microsoft Windows Server 2003 Standard Edition.
- Microsoft Windows Server 2003 Enterprise Edition.
- Red Hat Enterprise Linux 3.0 AS, ES, and WS, kernel level 2.4.21-15,
on X86 servers.
- SuSE Linux Enterprise Server 8.0 kernel level 2.4.21-231 (SP3) on X86
- IBM AIX V5.1, either 32-bit uniprocessor or 32-bit multiprocessor
with maintenance level 5.1 installed. The bos.mp (multiprocessor) or
bos.up (uniprocessor) package must be at level 126.96.36.199, or later. AIX
V5.1 32-bit HACMP environments are supported at the specified maintenance
- For IBM AIX V5.2, 32-bit or 64-bit, the bos.mp package must be at
level 188.8.131.52, or later for multiprocessor machines. The bos.up package
must be at level 184.108.40.206, or later for uniprocessor machines. AIX V5.2,
32-bit or 64-bit, HACMP environments are supported at the specified
- For IBM AIX 5.3, 32-bit or 64-bit, (to be completed text) AIX V5.3,
32-bit or 64-bit, HACMP environments are supported at the specified
- Sun Solaris 9, 64-bit, on Sparc servers.
The application servers, connected to the SAN File System, need either
single-pathing or multi-pathing device drivers to communicate with the
storage devices used with the SAN File System. A single-pathing device
driver provides the means for basic communication between a server and a
storage subsystem. A multi-pathing driver allows multiple fibre channel
paths to be connected to the storage subsystem and to be managed for
functions such as redundant path failover and load balancing. The
single-pathing or multi-pathing device drivers are the standard device
drivers for the supported storage subsystems.
Current software support summaries, including specific release and
firmware levels, are maintained at:
A Lightweight Directory Access Protocol (LDAP) directory must be
installed and available for configuration and use by the SAN File System
for storage administrator authentication and authorization. The LDAP
server or servers must be provided by the customer. Optionally, the
customer can install a LDAP directory on the server used for the Master
Back to top
Customers are responsible to provide the necessary prerequisite
hardware and software for the number of planned Metadata Servers and for
the Master Console required to implement the planned implementation of
the SAN File System V2.2 Software. In addition, planning for and
provision of the power and environmental support for the prerequisite SAN
File System hardware is a customer responsibility.
Planning considerations for SAN File System Software are listed in the
"IBM TotalStorage SAN File System: Planning Guide" GA27-4344.
V2.2 of the SAN File System has the limitation that files created by
Windows hosts can only be backed up and restored from a Windows host, and
files created by UNIX or Linux hosts can only be backed up and restored
from a UNIX or Linux host.
IBM Tivoli Storage Area Network Manager is included with the master
console software and has base support for up to 64 ports. Support for
more ports can be obtained by contacting IBM and licensing IBM Tivoli
Storage Area Network Manager.
Configuration guidelines and notices of temporary or unforeseen
problems, which can affect operations, are documented in the SAN File
System Software Release Notes, furnished publications for this product,
and/or as listed under configuration requirements and guidelines at:
The SAN File System is designed to provide performance that is
comparable to that of local file systems built on SAN-attached, high-
The SAN File System is a solution where all data read and written go
directly to the storage subsystem instead of through another file server,
and only metadata operations, such as create, delete, allocate, and look-
up go to the metadata server. Because of this architecture, most
applications are expected to see only a moderate difference in
performance from direct-attached storage while experiencing the benefits
that the SAN File System has to offer. Some I/O operations and workloads
that generate large amounts of metadata transactions will see a
difference in performance because of the distributed nature of the
system, but, in most cases, the performance is designed to be better than
a NAS-only solution. Application servers have a wide range of I/O
workloads and patterns, and therefore each may perform differently in
Direct storage access is just one of the advanced performance features
found in the SAN File System. Other performance features, included with
the SAN File System, are aggressive client side metadata caching, load
balancing through name space partitioning across the metadata servers,
and file striping across the volumes in a storage pool.
The installation services available are those for initial set-up and
configuration, including system set-up, client install, and definition of
policies and data migration as well as customer education.
In addition to the installation offering, there are two additional
services, "IBM Storage Strategy and Planning" and "IBM Technology
Assessment and Consulting Services for Storage Virtualization", that can
assist you in defining a storage virtualization strategy for your
Refer to the following Web site for additional information on
Back to top
Each new order of SAN File System software comes with publications.
The SAN File System Product Information CD-ROM and the Master Console
Publications CD-ROM contain an Information Center, displayable softcopy
publications as PDF files, and license information. The SAN File System
Product Information CD-ROM also contains the overview e-learning module
and managed object format (MOF) source code.
Back to top
Security, Auditability, and Control
The IBM TotalStorage SAN File System software administration interface
uses the security features of Microsoft Windows 2000 Server Edition, in
conjunction with Web Secure Shell (SSH) and Secure Socket Layer (SSL)
protocols and WebSphere.
The customer is responsible for evaluation, selection, and
implementation of security features, administrative procedures, and
appropriate controls in application systems and communication facilities.
(R), (TM), * Trademark or registered trademark of
International Business Machines Corporation.
** Company, product, or service name may be a trademark
or service mark of others.
Windows is a trademark of Microsoft Corporation.
UNIX is a registered trademark in the United States and other
countries licensed exclusively through X/Open Company Limited.
© IBM Corporation 2005.
Back to top