
533
High availability and load balancing using clusters
Cluster interactions with other security gateway features
Related information
For further information related to this topic, see the following:
■
■
“Adding or removing a cluster member”
■
“Adding or removing a cluster member”
■
“To remove a cluster member that has a redirected service”
■
“To add a security gateway with a redirected service to a cluster”
■
“Adding or removing a cluster member”
Modifying the RIP daemon for use with clusters
The routing information protocol (RIP) is an Interior Gateway Protocol (IGP) that helps you deploy
dynamic route discovery. By default, RIP advertises routes to other networks by giving the IP address
of the security gateway’s physical adapter to that network.
In a cluster environment, you can enable the RIP daemon to advertise the VIP address associated with
the interface so that HA/LB functions correctly. VIPs are used as next hop IP addresses in the routing
entries. You perform these configurations on all cluster members.
When using RIP within a cluster, the security gateway automatically creates RIP statements to the
virtual IP address. This way, the routing service is not disrupted in the event of a cluster node failure.
RIP is configured on a physical inside interface, but is applied using the VIP address.
Note:
OSPF will not advertise VIPs. Due to this, the failover time is between 30-40 seconds when OSPF
is running as a routing protocol for a cluster.
Using hot standby mode
The failover behavior of hot standby mode has ramifications for any activities that require you to
restart security gateways.
Two such occurrences are:
■
Failover and failback of VPN tunnels
■
Rebooting cluster members
Failover and failback of VPN tunnels
In hot standby mode, there is only one active cluster member. A second cluster member, the standby
cluster member, is available for failover. If the active cluster member becomes unavailable, the
standby cluster member becomes active.
If a VPN tunnel is connected through the active cluster member becomes unavailable, the VPN tunnel
moves to the standby cluster member.
When the active cluster member is restarted, the tunnels that are now connected to the standby cluster
member stop passing data. To enable the tunnels to pass traffic again, you must stop and restart the
standby cluster member. This forces the tunnels to return to the active cluster member.
Note:
The security gateway supports failover for the data channel (the channel that is used to transfer
files), but does not support failover for the control channel.
Содержание Security 5600 Series, Security 5400 Series,Clientless VPN 4400 Series
Страница 76: ...76 Managing administrative access Enabling SSH for command line access to the appliance...
Страница 242: ...242 Defining your security environment Controlling full application inspection of traffic...
Страница 243: ...243 Defining your security environment Controlling full application inspection of traffic...
Страница 269: ...268 Limiting user access Authenticating using Out Of Band Authentication OOBA...
Страница 319: ...318 Controlling traffic at the security gateway Blocking inappropriate content with content filtering...
Страница 373: ...372 Preventing attacks Enabling protection for logical network interfaces...
Страница 409: ...408 Providing remote access using VPN tunnels Multicast traffic through gateway to gateway IPsec tunnels...
Страница 509: ...508 Generating reports Upgrade reports...
Страница 553: ...552 Advanced system settings Configuring advanced options...
Страница 557: ...556 SSL server certificate management Installing a signed certificate...
Страница 861: ...860 Index...