Monday, April 5, 2010

EMC Celerra Character encoding for NFS

Enabling Unicode support for EMC Celerra

The Celerra Network Server supports clients in environments that use multibyte character sets. Multibyte character sets are supported by enabling universal character encoding standards (Unicode). The Celerra Network Server also supports the UTF-8 specification for character encoding, as required by NFSv4. For UTF-8 clients, file and directory names are stored as received, preserving the case of file and directory names, and no translation is required. For non UTF-8 clients, file system object names are translated to the UTF-8 specification.

Note: File systems that are not Unicode-enabled should be converted to Unicode before exporting them to NFSv4 clients. In addition, the Celerra character encoding method must be changed from the default to UTF-8.

Editing the xlt.cfg file

The xlt.cfg file defines how the Celerra Network Server should translate file and directory names for NFS and/or FTP clients into Unicode. It specifies the character mapping file(s) that should be used and allows you to configure all Data Movers to use either the same encoding or different encoding formats for different segments of your network.

NFSv4 requires UTF-8 for character encoding. By default, a Celerra system is configured to use ASCII Latin-1 character encoding. Before accessing data through NFSv4 clients, you must change the character encoding method to UTF-8 in the
/nas/site/locale/xlt.cfg file.

xlt.cfg file
The
xlt.cfg file is an ASCII text file. It contains a series of text strings, each consisting of up to six fields, that define the required client encodings to translate file/directory names into Unicode format.

The
xlt.cfg file is automatically installed in the /nas/site/locale directory during Celerra Network Server software installation. When configuring Celerra Network Server for Unicode, you issue the /nas/sbin/uc_config -update command to copy the contents of /nas/site/locale to
/.etc_common/xlt. This allows Data Movers to access xlt.cfg, if required, to complete the requested translations.

The format of text strings in xlt.cfg is as follows:
::

:::

uc_config command
uc_config
is a command for manipulating the translation files needed for Celerra Network Server’s international character support. These files are stored in a translation directory on a common file system, /.etc_common/xlt, that is available read-only to all Data Movers.

The translation directory is created when Celerra Network Server software is installed on the Control Station. If this directory is destroyed and needs to be re-created, you can do so by running the uc_config command with the -setup option.

To check for the existence of the common file system and translation subdirectory and create it if necessary, type:

Example 1:
$ /nas/sbin/
uc_config -setup

If you receive the following message, your setup is correct:
Common Unicode translation subdirectory already exists.

If you do not receive this message, contact EMC Customer Service for assistance.

Example 2:
To list the set of translation files available for conversion/update, type:

$ /nas/sbin/uc_config -list

8859-1.txt big5.txt eucjp1.txt unidata2.txt cp437.txt sjis.txt xlt.cfg

Example #3
To update all configuration files in
/.etc_common/xlt, type:

$ /nas/sbin/uc_config -update

operation in progress (not interruptible)...

done

Example #8
To confirm that the configuration of
xlt.cfg is correct for a client with an IP address of 168.159.30.77, for server_3, type:

$ /nas/sbin/uc_config -verify 168.159.30.77 -mover server_3
server_3:
168.159.30.77 is UTF-8

No comments: