E.0 Comparison of NSS on Linux and NCP Volumes on Linux POSIX File Systems

This section compares features and capabilities of the Novell Storage Services file system on Novell Open Enterprise Server 2015 SP1 to those of NCP volumes on Linux POSIX file systems such as Ext3, XFS, BtrFS, and Reiser. For information, see Managing NCP Volumes in the OES 2015 SP1: NCP Server for Linux Administration Guide.

For information to help you choose from among the numerous Linux file system offerings, see Overview of Linux POSIX File Systems in the OES 2015 SP1: Linux POSIX Volume Administration Guide.

Feature Description

NSS on OES 2015 and Later

NCP Volumes on Linux POSIX File Systems

Management interfaces

Novell iManager Storage plug-in. For information, see Section 10.1, Novell iManager and Storage-Related Plug-Ins.

NSSMU for Linux

NLVM

NSS utilities in a terminal console

NSS commands in the NSS Console (NSSCON)

Novell Remote Manager for Linux (browse only)

YaST > Partitioner for managing devices

Novell Remote Manager for Linux (Managing Shares)

NCP commands in the NCP Console (NCPCON)

Various Linux commands and utilities in a terminal console.

NSSMU

NLVM commands

File system trustees and trustee rights to control access to directories and files

Yes, works with or without concurrent running of NCP Server.

Yes, requires NCP Server to enforce the rights and access on the extended attributes.

File access protocols

NCP

CIFS/Samba using Novell Samba

Linux NFS (version 3)

Linux NFS and Samba requires users to be Linux-enabled with Linux User Management. The service must also be LUM enabled.

Novell AFP for Linux

Novell CIFS for Linux

NCP

CIFS/Samba using Novell Samba

NCP Volumes and Samba requires users to be Linux-enabled with Linux User Management. The service must also be LUM enabled.

File system directory and file attributes to control functions available for directories and files

Files and Folders plug-in to iManager

Novell NetStorage

Novell Client

Novell Remote Manager for Linux. See Viewing Key NSS Directory and File Attributes as Linux POSIX Permissions in the OES 2015 SP1: File Systems Management Guide.

Not applicable. Use POSIX file and directory attributes.

Directory quotas

Yes, requires the Directory Quotas attribute to be enabled.

Yes

User space quotas (user space restrictions)

Yes, for OES Linux SP2 and later

Yes, if the Linux file system being used under the NCP share supports user quotas and the Linux file system resides on a local, iSCSI, or Fibre Channel drive. All users of the NCP volume must be LUM enabled. Manage the user quotas using the Linux file system tools.

Default mount location for NSS pools

/opt/novell/nss/mnt/.pools/poolname

Not applicable.

Volume name space

Long is the default name space, which is case insensitive.

You can specify the UNIX name spaces on mounting the NSS volume to make its directory names and filenames case sensitive. Using UNIX name space slows performance compared to using Long.

For example:

ncpcon mount /opt=ns=<long|unix|dos|mac> <volume_name> 

The name space options are case sensitive.

UNIX; no support for case insensitive names.

Salvage for deleted volumes, directories, and files

Yes

No

Volume encryption

Yes, for OES SP2 and later

Yes, for Reiser

File compression

Yes

No

Data shredding (secure deletion)

Yes, up to 7 times

No

Online resizing of volumes and pools

Yes

Yes, depending on the file system

Multiple I/O paths to storage media

For information, see Managing Multipath I/O for Devices in the SLES 11: Storage Administration Guide.

No; NSS-specific multipath I/O tools as are not available on Linux.

Use the Linux Device Mapper driver support for mutlipath I/O on devices where you plan to create NSS file systems.

Use the Linux Device Mapper driver support for mutlipath I/O on devices. (NCP is not required to make this work.)

Software RAID support

RAID 0, 1, 5, 0+1, and 5+1.

RAID 0, 1, 4, 5 and 6. RAID 0+1 can be created using the Linux mdadm(8) command as a complex RAID using the RAID 0+1 option, or as a nested RAID.

Pool snapshot

(retain point-in-time version of a pool using block-level copy on write)

Yes, using iManager, NSSMU, or NLVM command line interface.

Snapshots of cluster-enabled pools is not supported.

Depends on the file system. NLVM supports device snapshots for the devices it manages. (NCP is not required to make this work.)

Hard links

Yes; enhanced hard links support is available in OES 2 and later.

For information, see Section 25.0, Managing Hard Links.

Yes

Backup support

Yes, using Novell Storage Management Services for Linux.

For information, see Section 27.0, Managing Backup and Restore for Data and Trustee Information.

Yes. You can use SMS. For more information, see POSIX File System Support.

Data migration from NSS volumes on NetWare

Yes

Yes

Novell Archive and Version Services

No

No

Novell Distributed File Services

For information, see the OES 2015 SP1: Novell Distributed File Services Administration Guide for Linux.

Yes, for OES 2 and later.

NSS volumes on OES 2015 can contain junctions or be a junction target. NSS volumes on OES 1 can be a junction target, but junctions are not supported in the volume.

Only as targets of junctions in OES 2 and later.

DFS does not support junctions on NCP volumes.

Dynamic Storage Technology

For information, see the OES 2015 SP1: Dynamic Storage Technology Administration Guide.

Yes

Not available.

Novell Cluster Services for Linux

For information, see the OES 2015 SP1: Novell Cluster Services for Linux Administration Guide.

Yes

For information, see Creating Cluster-Enabled Pools and Volumes.

Yes; cluster the Linux POSIX file system, then create the NCP volume on it. For information, see Upgrading and Managing Cluster Resources for Linux POSIX Volumes with CSM Containers.

You can NCP enable the clustered Linux volume as you create it by using NSSMU or the 'nlvm create linux volume' command. For information, see Clustering LVM Volume Groups with Novell Cluster Services in the OES 2015 SP1: Linux POSIX Volume Administration Guide.

Creating a Linux Volume on a Device that Contains a Novell Partition in the OES 2015 SP1: NLVM Reference.

Auditing

Yes

No

Novell Transaction Tracking System (TTS)

No

Use the Journal mode for Linux POSIX file systems that support journaling.

Operating system version detection

Default process

Default process

Device maintenance support

Activate and deactivate devices by pool.

Activate and deactivate devices using Linux tools.

Cache balancing for NSS cache buffers

You can specify a minimum cache buffer size.

For information, see Tuning NSS Performance.

Integrated with the Linux file system cache.

CD and DVD device recognition

No; not managed by NSS.

Use Linux services to mount CDs and DVDs as Linux volumes.

Yes, default

Ability to access DOS partitions as on a NetWare server

No; not managed by NSS. Use Linux services instead.

Yes, using Linux services.

Default mount location for NSS volumes

/media/nss/volumename

Not applicable.

Interface

64-bit

64-bit

Character format

Unicode

UTF-8

Maximum device size recognized (physical or logical)

8 TB for NSS32 pools and 8 EB for NSS64 pools.

NSS management tools recognize devices up to 8 EB, and support both the GPT and DOS partitioning schemes. For use in an NSS pool, the maximum device sizes supported are:

8 TB (GPT) for NSS32 Pools

8 EB (GPT) for NSS64 Pools

2 TB (DOS) for NSS32 and NSS64 pools

For a 64-bit OS:

  • 2 to 32 TB for Ext2 or Ext3, depending on the block size

  • 16 TB (minus 1 Byte) for Reiser

  • 16 EB for Btrfs and XFS

Maximum software RAID size (combined total of all member segments)

8 TB (GPT) for NSS32 Pools

8 EB (GPT) for NSS64 Pools

2 TB (DOS) for NSS32 and NSS64 pools

NOTE:The pools are currently limited to 8 TB for NSS32 and 8 EB for NSS64, RAID1 sizes for pool objects are also limited to 8 TB or 8 EB depending on the pool type.

See Maximum device size recognized.

Minimum software RAID segment size

12 MB per segment

Depends on the file system.

Maximum partition size

8 TB for NSS32 and 8 EB for NSS64 pools

Valid Range: 12 MB to 8 TB for NSS32 pools, and 12 MB to 8 EB for NSS64 pools

NOTE:The Maximum segment size corresponds only to GPT partition. If all devices are partitioned using DOS, the size limits to 2 TB.

Up to 16 TB or 16 EB, depending on the file system and block size as noted above.

Maximum number of partitions (logical or physical devices) per pool

No practical limit

Not applicable.

Maximum pool size

8 TB for NSS32 pools (using 4 or more partitions of up to 2 TB each and if device is partitioned with GPT which is of >=8 TB size you can have a pool of 8 TB with single partition only)

8 EB for NSS64 pools

Up to the partition size, depending on the file system.

Minimum pool size

12 MB for both NSS32 and NSS64 pools.

Not applicable.

Maximum size of a volume

Up to 8 TB or 8 EB, depending on the pool size, pool type and available space in the pool.

Volume quotas can be overbooked. For information, see Section 19.2, Guidelines for NSS Volumes.

Up to the partition size, depending on the file system.

Maximum file size

Up to 8 TB or 8 EB, depending on the volume size, pool type, and available space in the volume.

2 GB to 2 TB for Ext2 or Ext3, depending on the block size.

Up to 8 TB for Reiser.

Up to 8 EB (minus 1 Byte) for XFS

Up to 8 EB for Btrfs (due to Linux kernel limit)

Maximum number of files per volume

(In practice, how many files be managed is limited only by the file browser’s and application’s ability to list and access the files.)

Up to 8 trillion (10E12), regardless of how many name spaces are loaded.

Up to 4 billion (10E9) files in a single directory.

Up to 8 trillion (10E12), regardless of how many name spaces are loaded.

Maximum number of files open concurrently

1 million (10E6)

Millions (10E6), depending on the file system

Maximum number of volumes per server

Unlimited NSS data volumes, but only 255 can be mounted at a time

Unlimited

Time to mount a volume

Requires only a few seconds

NSS uses a journaling file system and does not need to scan the entire file system to create a directory entry table (DET) and to load a File Allocation Table (FAT).

Depends on the file system; from a few seconds to a few minutes.

Time to repair corrupted volume

Up to several hours, depending on the volume size.

Up to several hours, depending on the volume size