Enabling CS Linux servers

This section describes how to enable CS Linux on a computer that was installed as a server (that is, with the SNA node components installed). If you are enabling CS Linux on a client, see Enabling and disabling Remote API Clients on AIX or Linux.

You must enable CS Linux on the local system before you can configure or manage the local node (either locally or from a remote CS Linux node).

To enable the CS Linux software, enter the following command at the Linux command prompt:

sna start [ -s ] [
-m
kernel_memory_limit] [
-t ]
Note: When you use the sna start command, the CS Linux software uses the directory from which you issued the command as its current working directory, and maintains one or more open file descriptors in that directory. This means that you will not be able to unmount the file system containing that directory while the CS Linux software is running. To avoid problems, you should start the CS Linux software from a directory on a filesystem that does not need to be unmounted; for example, you could use cd / to change to the root directory before using the sna start command.

When you install CS Linux, the installation utility automatically updates the startup file /etc/rc.d/init.d/snastart to include the sna start command. This ensures that CS Linux is started automatically at system startup. If you do not want CS Linux to be started automatically, you can remove or comment out this line, and then follow the instructions in this section to enable the CS Linux software manually.

The parameters and options for the sna start command are as follows:

-s
Specifies that CS Linux should not write messages to the system console. If you do not use this option, CS Linux writes messages to the console when it ends, and also writes the text of certain error log messages to the console as well as to the log file.
-m kernel_memory_limit
Specifies the maximum amount of kernel memory, in kilobytes, that CS Linux should use at any time. (Kernel memory is used for internal data structures.) If a component of CS Linux attempts to allocate kernel memory that would cause the total amount of memory currently allocated to CS Linux components to exceed this limit, the allocation attempt fails.

If you do not use this option, kernel memory usage is not limited.

-t
Activates tracing on all interfaces between kernel components, and also client/server tracing. (This option does not turn on DLC tracing.) Tracing enables you to diagnose problems that occur during startup. If you do not use this option, tracing is inactive at all interfaces; you can then activate it on specific interfaces as required, using the command-line administration program snaadmin.

Tracing on all interfaces degrades the performance of CS Linux components. After the software is enabled, you can use the command-line administration program snaadmin to stop tracing on any interfaces where it is not required. For more information about tracing, refer to IBM Communications Server for Data Center Deployment on Linux Diagnostics Guide.

CS Linux writes messages to standard error (normally your terminal's screen) to indicate that it is initializing, and to indicate whether initialization completes successfully.

If initialization fails, the messages include information about the cause of the error, and (where appropriate) additional information such as the Linux operating system error message. The text written to standard error may also include a message indicating that you can find further information in the error log file. The sna start command then ends with a nonzero exit code that indicates the nature of the error.

For more information about exit code values, refer to IBM Communications Server for Data Center Deployment on Linux Diagnostics Guide.