HP 5500 EI Series Configuration Manual page 189

to all the switches within specified hops.
To avoid concurrent responses to an NTDP topology-collection request causing congestion and deny of
service on the management switch, a delay mechanism was introduced. You configure the delay
parameters for NTDP on the management switch. As a result:
Each requested switch waits for a period of time before forwarding an NTDP topology-collection
•
request on the first NTDP-enabled port.
After the first NTDP-enabled port forwards the request, all other NTDP-enabled ports on the
•
requested switch forward the request in turn at a specific interval.
Cluster management maintenance
1. Adding a candidate switch to a cluster:
You should specify the management switch before creating a cluster. The management switch
discovers and defines a candidate switch through NDP and NTDP protocols. The candidate switch
can be automatically or manually added to the cluster.
After the candidate switch is added to the cluster, it can obtain the member number assigned by
the management switch and the private IP address used for cluster management.
2. Communication within a cluster:
In a cluster the management switch communicates with its member switches by sending handshake
packets to maintain connection between them. The management/member switch state change is
shown in
Figure 59 Management/member switch state change
A cluster manages the state of its member switches as follows:
• After a candidate switch is added to the cluster and becomes a member switch, the management
switch saves its state information and identifies it as Active. The member switch also saves its state
information and identifies itself as Active.
The management switch and member switches send handshake packets. Upon receiving the
•
handshake packets, the management switch or a member switch keeps its state as Active without
Figure 59.
179