Using SAMBA Toolkit in a ServiceGuard for Linux (SG/LX) Cluster Version A.01.00, June 2003

Using SAMBA Toolkit in a ServiceGuard for Linux (SG/LX) Cluster Version A.01.00
Setting Up SAMBA Server Application for a SG/LX Cluster
Chapter 116
Toolkit Features
In order to support multiple SAMBA instances running on a SG/LX
cluster, you need to define SAMBA common and specific option values for
a specific SAMBA package. These options are used as command
parameters for starting SAMBA daemons (smbd and nmbd). (Each
ServiceGuard package represents a single SAMBA instance in the
cluster. You can create several ServiceGuard SAMBA packages to make
multiple instances co-existing in the cluster.)
Two variable values you need to set as follows:
- SAMBA Base Configuration File Path of common configuration file for
all SAMBA packages in the cluster: Each SAMBA instance must have
this base configuration file that contains SAMBA server configuration
data for all SAMBA instances in the cluster.
- Specific Configuration File Path of a SAMBA server instance: This file
contains configuration data for a specific SAMBA instance in addition to
the data provided in the Base Configuration File. (The SAMBA daemons
will load the configuration data from both configuration files when they
are started.)
The monitoring process for a SAMBA server will be launched as a SG/LX
Service. You need to provide Service information in order to make the
SG/LX to launch the monitoring process and interact with the SG/LX.
Five variable values need to be provided user as below:
• Name of a SG/LX Service: This name should be identical to the
service name defined and registered in SG/LX package configuration.
• Command to be used for the SG/LX Service: This will be the SAMBA
toolkit monitor script name and its parameters.
• Restart option for the SG/LX Service: For Example “-r 2” means retry
2 times for launching a service.
• Monitoring interval: giving the information to the monitor process of
how often the process needs to check the health of the running server
daemons.
• Retry times of monitoring process: retry times of verifying the fail of
server daemons.