Whether IP addresses that have been planned during the first deployment of OceanStor 9000 need to be re-planned for follow-up expansion

Loose and strict explicit paths are usually used in TE path planning. When configuring an explicit path, you can define a node that traffic must pass through or a node that traffic must bypass. Parameter include means that an LSP must pass through a specified node. Parameter exclude means that an LSP cannot pass through a specified node. When configuring a strict explicit path, you must specify the next hop and ensure that the next hop is directly connected to its upstream node. A loose explicit path allows other nodes to exist between the source and sink. In the IP RAN solution, loose explicit paths are usually used to facilitate future node addition or deletion. When planning an explicit path for an LSP, note the following rules: 1. The explicit path must not pass through the destination device. In other words, if you have specified a node that an explicit path must pass through, then the explicit path must pass through this node before reaching the sink. 2. The explicit path must not pass through a path twice. 3. Among multiple paths between two nodes that an LSP must pass through, the explicit path uses the one with the lowest IGP cost. If you do not specify any node for the explicit path to pass through, the explicit path uses the path with the lowest IGP cost between the source and sink. 4. When overlapping is enabled, the primary and secondary paths try their best to avoid sharing the same path, but may share the same node. It is recommended that you use the outbound interface IP address of the source or sink for route restriction. If you want to exclude multiple paths from an LSP, specify the intermediate nodes that need to be excluded besides specifying the outbound interface IP address of the source or sink. When configuring explicit paths, you are also advised to configure overlapping, route re-optimization, and automatic hot-standby backup path computation.

For details about the planning of IP addresses on the OceanStor 9000 V100R001C01 10GE network, see OceanStor 9000 V100R001C01 Planning Guide.

For details about the planning of IP addresses on the OceanStor 9000 V100R001C01 InfiniBand network, see OceanStor 9000 V100R001C01 Planning Guide.

Yes, but you are advised to plan consistent IP addresses for the IPMI interface of each node for the ease of remote node management. In addition, you are advised to plan all IP addresses in a management network segment that is usually a private network segment. The private network segment has no requirements on the number of IP addresses and allows requests for application.

Management ports in OceanStor 9000 V100R001C01 include NIC0 and IPMI ports. NIC0 ports are used for logging in to system management software, as well as configuring, maintaining, and monitoring the system. IPMI ports on each node are used for hardware system management, including viewing hardware component status and remotely powering on or off nodes. The entire system uses only NIC0 ports on the first three nodes, and one floating management IP address (floating on the NIC0 ports on the three nodes) is planned. An IP address is separately planned for the IPMI port on each node.