If you are wondering how to use the EP3 administrative interface, this article will give you a few tips and strategies. The first step is to export the EP3 XML data dump from your hosting provider. After exporting the XML file, please navigate the EPrints administration page and upload it. Confirm each step of the import process. You can re-use the exported files if you wish. Once the import process has begun, you can check the status of your imported files in the administration interface.
BD multicast address
The BD multicast address is unique to every BD. A BD can receive a packet only from another BD with the same multicast address. The spine will flood the box to the leaves of the part of its local Pod. A BD uses its multicast address to communicate with leaves in its IP subnet. A small leaf receives a multicast packet encapsulated in a VXLAN packet. A BD can learn about this multicast address by sending a Multi-Pod-LL TEP.
The ARP response sent by an endpoint connected to a small leaf is forwarded to the remote leaf. The EP3 then sends a corresponding unicast ARP response with its destination MAC address as the EP1 and its sender MAC as the EP3 leaf. It also learns about the endpoint connected to the small leaf via the spine. The spine will then pass the packet back to the endpoint.
COOP information in SW DB
The small leaf will attempt to build a COOP or spine proxy connection during spine connectivity failure and forward traffic based on data plane activity. Whenever a packet arrives at an alternate leaf, the leaf will first check its COOP or spine database for the destination IP address and then forward it to the local leaf. This procedure prevents the EP from sending traffic to the wrong destination when the spine fails.
The EP3 sends an ARP request to its leaf during the ARP process. When it receives the request, it encapsulates it in an ARP packet and forwards it to the leaf connected to the small leaf. The leaf then forwards the packet with a destination IP address of RL-vPC-TEP and RL-Ucast-TEP.
Reachability between spines and remote leaves
The Remote Leaf builds its local software database using the COOP database of the spines. This local software database contains information about all endpoints in the VRFs. For example, if Remote location 1 has deployed VRF1 and remote location 2 has deployed VRF2, the RL will build the local software database based on the COOP database of the spine. When a small leaf encounters a problem with the spine, it can use the local SW DB to forward the packet.
If the spines are logically connected, they can reach remote leaves using a VXLAN multicast address. In multi-pod environments, the spines use APIC controllers distributed among the Pods. The RLs can still manage the small leaf. However, the spines are not able to reach small leaves directly. As a result, the spines can’t HREP L2 multicast traffic directly to small leaves.
Tips and strategies for using the EP3 administrative interface
While configuring your EP3 server, you may not realize how much sensitive data you send across the administrative interface. To ensure security, configure the administration interface only to allow authorized users. If remote administration is necessary, you can restrict the users and use encrypted channels. To further reduce risk, enforce strong authentication on all users and use IPSec policies for all connections. Then, make sure that your administrators authenticate themselves through solid authentication.