Symantec Security 5600 Series, Security 5400 Series,Clientless VPN 4400 Series, Руководство администратора

Страница: 524 / 861

523 High availability and load balancing using clusters
Managing clusters
Rebooting a cluster
You may need to reboot for several reasons. For example, when you add or remove an interface, when a
cluster member is not responding, after you dissolve a cluster, when you enable hot standby, or when a
hotfix has been applied.
When the cluster reboots, it reboots only half of the cluster members at a time. For example, if you
have a five member cluster, two members reboot first, followed by the next two members, followed by
the cluster member from which you are negotiating the process. This ensures that the cluster is always
available.
The default setting for reboot timeout is 120 seconds. You can change the wait-time value if it is too
low by increasing the parameter half.reboot_wait_time in Advanced Parameters.
Verify that cluster configurations have been propagated, by viewing the cluster status.
Note: This stepped reboot process does not apply to rebooting when you add delete an interface or a
new cluster member. When you reboot for these two situations, all cluster members reboot
simultaneously.
Prerequisites
None.
To reboot a cluster
1 In the SGMI, on the Tools menu, click Cluster > Reboot Cluster .
2
In the Confirmation panel, review the cluster members that you want to reboot, and then click
Finish .
3 If a cluster member does not reboot, do one of the following:
■
To wait longer for the cluster members to reboot, click OK .
This allows another 120 seconds to reboot.
■
To exit the reboot process, click Cancel .
You can fix the problem and restart the reboot process.
Once the process has completed successfully, a message displays saying the cluster member will
reboot in 8 seconds.
4 Do one of the following:
■
To reboot immediately, click OK .
■
Otherwise, to reboot later, click Close .
Related information
For further information related to this topic, see the following:
■
“Viewing the cluster status in the SGMI” on page 528
■
“Confirmation panel” on page 805