SUSE Linux Enterprise 10
Tech Specs
File System Support
ReiserFS v3, ext3, XFS, and OCFS2 Supported
The SUSE® Linux Enterprise platform is the only Linux distribution that offers broad support for a number of file systems out of the box. In SUSE Linux Enterprise 10, Novell is providing enterprise support for the ReiserFS v3, ext3, XFS and OCFS2 file systems. Each of these file systems has its strengths for specific application and workload types. Offering this broad support allows customers the opportunity to deploy the file system that best meets the needs of their preferred data storage application or service.
The Right File System for Specific Tasks
In SUSE Linux Enterprise 10 and prior versions, the best file system for scalability and all-around performance has been ReiserFS v3. For more specific uses, ext3 and XFS have been recommended, as well as OCFS2 in SUSE Linux Enterprise Server 9 SP2 and the latest SUSE Linux Enterprise 10 release for Oracle RAC workloads. ext3 has been recommended for small file systems typically used in edge services in prior distributions. XFS has been recommended for large file systems that also contain large files (such as video rendering). All of these file systems will continue to be supported well into the future.
File System Future Directions
In open source development, it is expected that changes be communicated early. For SUSE Linux Enterprise 11, Novell will be changing the default file system in new installations from ReiserFS 3 to ext3. The openSUSE® builds has begun to reflect this change in openSUSE 10.2. These are the community builds that are on the path to the next enterprise distribution release: SUSE Linux Enterprise 11.
Kernel Limits
This table summarizes the various limits which exist in our recent kernels and utilities (if related) for SUSE Linux Enterprise 10 Service Pack 2. This is Novell's award-winning family of Linux operating system products based on a common code base, including SUSE Linux Enterprise Server 10, SUSE Linux Enterprise Desktop 10, SUSE Linux Enterprise Real Time 10 and others.
φ = insufficient data| SLE 10 SP2 (2.6.16.60 |
x86 (IA-32) |
ia64 (Itanium) |
x86_64 (AMD64/EM64T) |
s390x (IBM System z) |
ppc64 (IBM System p) |
|
|---|---|---|---|---|---|---|
| Kernel related items | ||||||
| CPU bits | 32 | 64 | 64 | 64 | 64 | |
| max. # CPUs | 32 (up to 128 with bigsmp kernel on certified systems) |
up to 4096 (on certified SGI systems) |
32 (up to 128 on certified systems) |
64 | 128 | |
| max. RAM (theoretical / certified) | 64/16 GiB | 1 PiB/4 TiB | 64 TiB/512 GiB | 4 TiB/256 GiB | 1 PiB/512 GiB | |
| max. swap space | up to 32 * 64 GB | |||||
| max. user-/kernelspace | 3/1 GiB | 2 EiB/φ | 128 TiB/128 TiB | φ/φ | 2 TiB/2 EiB | |
| max. #processes | 1048576 | |||||
| max. #threads per process | tested with more than 120000; maximum limit depends on memory and other parameters | |||||
| max. size per block device | up to 16 TiB on 32-bit architectures, and up to 8 EiB on 64-bit architectures | |||||
| Filesystem related items | ||||||
| max. filesystem size | ext2/3: 16 TiB (244 bytes) nfs v2: 8 EiB (263 bytes) nfs v3: 8 EiB (263 bytes) nfs v4: 8 EiB (263 bytes) reiserfs: 16 TiB (244 bytes) xfs: 8 EiB (263 bytes) ocfs2: 16 TiB (244 bytes) |
|||||
| max. file size | ext2/3: 2 TiB (241 bytes) jfs: 8 EiB (263 bytes) nfs v2: 2 GiB (231 bytes) nfs v3: 8 EiB (263 bytes) reiserfs: 1 EiB (260 bytes), but page cache limits this to 8 TiB on architectures with 32 bit int xfs: 8 EiB (263 bytes) ocfs2: 1 EiB (260 bytes) |
|||||
Remarks:
- The maximum file size above can be larger than the filesystem's actual size due to usage of sparse blocks. It should also be noted that unless a filesystem comes with large file support (LFS), the maximum file size on a 32-bit system is 2 GiB (231 bytes). Currently all of our standard filesystems (including ext3 and ReiserFS) have LFS, which gives a maximum file size of 263 bytes in theory. The numbers given in the above table assume that the filesystems are using 4 KiB block size. When using different block sizes, the results are different, but 4 KiB reflects the most common standard.
- The rationale for the use of theoretical vs. certified in the following tables is as follows: theoretical means that's what should work, at least theoretically, while certified describes something which Novell and our Partners have tested and certified to work on existing machines in real life scenarios. With this explanation, the values for the x86, i.e. 32-bit bigsmp kernels should have been 64 GiB in the first place, but everyone knows that would just be too theoretical as it doesn't really work (and if it did in older kernels, only for very special configurations and workloads).
- In this document: 1024 Bytes = 1 KiB; 1024 KiB = 1 MiB; 1024 MiB = 1 GiB; 1024 GiB = 1 TiB; 1024 TiB = 1 PiB; 1024 PiB = 1 EiB (see also http://physics.nist.gov/cuu/Units/binary.html)
System Requirements
The following list of resources includes links to each Tech Specs page for the SUSE Linux Enterprise offerings—both within the Server and Desktop product groups. The individual Tech Specs pages include information on product System Requirements and supported Processor Platforms. Also mentioned in the list below are resources on system requirements for SUSE Linux Enterprise Partner Certification.
Server Products
- SUSE Linux Enterprise Server
- SUSE Linux Enterprise Real Time
- Open Enterprise Server
- SUSE Linux Enterprise Virtual Machine Driver Pack
- SUSE Linux Enterprise Server Priority Support for SAP
Desktop Products
- SUSE Linux Enterprise Desktop
- SUSE Linux Enterprise Thin Client
- SUSE Linux Enterprise Point of Service
Partner Certification
Accessibility
As a leading software company, Novell believes that it is the responsibility of vendors to bring the benefits of technology to people with disabilities. As part of our engineering process we integrate accessibility technologies into our products. We believe that universal access to computers is an important issue.
As part of this commitment to accessibility, Novell has contributed to open accessibility, in conjunction with Sun and IBM, over a long period. Our work is extensive ranging from infrastructure, through debugging and test to fast, native OpenOffice.org accessibility.
Recently, Novell extended its technical collaboration with Microsoft to make it easier for all software companies and developers to create and deliver accessible applications across both Windows and Linux platforms, which will improve access to computer technology for people with disabilities even further.
As part of this new collaboration Novell will develop and deliver an adapter that allows applications written using Microsoft's User Interface Automation (UIA) API to work well with existing Linux accessibility projects and complement the investments made by Sun Microsystems, IBM Corp., and others. To promote interoperability between leading accessibility frameworks in the market, Novell's work will be open source and will make the UIA framework cross-platform building on the Linux Accessibility Toolkit (ATK), shipping in SUSE Linux Enterprise, Red Hat Enterprise Linux and Ubuntu Linux. The UIA solution will ensure interoperability of nonvisual access to important vertical apps being migrated from Windows to Linux.