The ITD health
monitoring module periodically monitors nodes to detect any failure and to
handle failure scenarios.
ICMP, TCP, UDP, DNS and HTTP probes are supported to probe each node periodically for health monitoring. A probe can be configured
at the device-group level or at node-level. A probe configured at the device-group level is sent to each node member of the
device-group. A probe configured at a node-level is sent only to the node it is associated with. If a node-specific probe
is configured, only that probe is sent to the node. For all the nodes that do not have node-specific probe configuration,
the device-group level probe (if configured) is sent.
![](https://www.cisco.com/content/dam/en/us/td/i/templates/note.gif) Note |
HTTPS probe is not supported on ITD.
|
IPv4 Control Probe for IPv6 Data Nodes
For an IPv6 node (in an IPv6 device-group), if the node is a dual-homed
node (that is, it supports IPv4 and IPv6 network interfaces), an IPv4 probe can
be configured to monitor the health. Since IPv6 probes are not supported, this
provides a way to monitor health of IPv6 data nodes using a IPv4 probe.
![](https://www.cisco.com/content/dam/en/us/td/i/templates/note.gif) Note |
IPv6 probes are not
supported.
|
Health of an Interface Connected to a Node
ITD leverages the IP
service level agreement (IP SLA) feature to periodically probe each node. The
probes are sent at a one second frequency and sent simultaneously to all nodes.
You can configure the probe as part of the cluster group configuration. A probe
is declared to have failed after retrying three times.
Node Failure Handling
Upon marking a node as
down, the ITD performs the following tasks automatically to minimize traffic
disruption and to redistribute the traffic to remaining operational nodes:
-
Determines if a
standby node is configured to take over from the failed node.
-
Identifies the
node as a candidate node for traffic handling, if the standby node is
operational.
-
Redefines the
standby node as active for traffic handling, if an operational standby node is
available.
-
Programs
automatically to reassign traffic from the failed node to the newly active
standby node.