This section describes how to use sybcluster to configure and start the cluster. Make sure you first follow the instructions in:
You can view a complete sybcluster session that includes answers to all the prompts in the Appendix C, “A sybcluster Sample Session.”
Use sybcluster to create and configure
a shared-disk cluster. To create logical clusters, use
the Adaptive Server plug-in or system stored procedures described
in the “Managing the Workload” chapter of the Users
Guide to Clusters.
Start sybcluster. For example:
sybcluster -U uafadmin -P -C mycluster -F "blade1,blade2,blade3"
See the Users Guide to Clusters for complete syntax and usage information for sybcluster.
Enter:
create cluster
sybcluster prompts you for all necessary information, one parameter at a time. If there is a default value, sybcluster displays it in the command prompt. To accept the default, press Enter. Otherwise, enter the correct value, then press Enter.
sybcluster prompts for the following. See “Worksheet for creating a cluster”.
The name of the cluster to create if you did not set the default cluster on the command line.
The maximum number of instances to create for the cluster.
The number of nodes in the cluster.
sybcluster displays a list of available host machines. In this list are all nodes specified by the -F and -d parameters on the sybcluster command line with a configured and running Unified Agent as set up in “Setting the environment and starting the Unified Agents”. Select nodes from this list.
You can create a simulated cluster by hosting all instances
on a single node. For best performance, make sure that the total
number of engines across all instances running on the single node
does not exceed the number of CPUs, or cores, on the host machine.
Additionally, the default setting for the runnable process search count parameter is '3' . Sybase recommends using this default setting in this environment. This can be set using the sp_configure stored procedure.
The number representing the cluster node. sybcluster numbers the remaining nodes in the cluster starting with this number.
The full path for the quorum device. For example, /dev/raw/raw11.
Trace flags.
The full path for the master device. For example, /dev/raw/raw12.
The size of the master device.
The size of the master database.
The page size of the master database, in kilobytes.
The full path to the system procedure database device, sysprocsdev. For example, /dev/raw/raw13.
The size of the system procedure database device.
The size of the system procedure database.
The full path to the Sybase system database device, systemdbdev. For example, /dev/raw/raw14.
The size of the system database device.
The size of the system database.
Does this cluster have a secondary network? (Y).
If you enter “Y” to a secondary network, sybcluster uses the default port number of 15100 as the starting port number for the range of port numbers required so that the instances can exchange messages, and reserves that number of ports, starting with the default value.
Adaptive Server uses several sockets for each interconnection.
The number of ports required for each instance is 5 multiplied by
the maximum number of instances.
If you enter “N” to a secondary network, sybcluster asks for the beginning port number, and calculates the number of additional ports required and reserves that number of ports. The default value is 15100.
The $SYBASE home directory. For example, /remote/var/sybase.
The full path to the “.sh” environment shell script. For example, /remote/var/sybase/SYBASE.sh.
The Adaptive Server home directory. For example, /remote/var/sybase/ASE-15_0.
Ensure that the interfaces file does not already
contain the cluster or instance information. sybcluster adds
this information during the Create Cluster process
The directory containing the interfaces file. For example, /remote/var/sybase. sybcluster adds the correct cluster and instance information during configuration.
The full path to the dataserver configuration file. For example, /remote/var/sybase/mycluster.cfg. sybcluster must be able to find this file during configuration.
sybcluster displays the node name and prompts for values for each instance, one node at a time.
Instance name.
Query port number for the instance. Ensure that it is available and not in use by other applications.
The primary protocol address for the instance. For example, 10.0.1.1.
The secondary protocol address for the instance (if you answered Y to a secondary network). For example, 10.0.1.2.
Local system temporary database device. The name of the Adaptive Server database device for the local system temporary (LST) database. You must create local system temporary database on a shared disk.
The LST device path. For example, /dev/raw/raw15.
The LST device size.
If you place all LST databases on the same device, the
device size must be large enough for all LST databases.
The LST database name.
The LST database size.
sybcluster asks if you want to add another instance. If you enter Y, sybcluster repeats step 4 for the next instance.
sybcluster asks if you want to save the cluster configuration you have entered.
If you enter Y, sybcluster stores the configuration in an XML file that you can edit and replay using the sybcluster command:
create cluster cluster_name file file_name
sybcluster asks if you want to create the cluster.
If you enter Y, sybcluster configures the cluster using the information you have entered. This may take several minutes.
While creating the cluster, sybcluster asks if you want to check if each cluster device supports IO fencing. If you answer yes, sybcluster checks all devices and reports the IO fencing capability of each one.
Configuring the cluster