Worksheet for verifying network requirements
For each cluster, verify the following network requirements.
Every node on the cluster must have three IP addresses that are reserved in advance. Most hardware platforms require two IP addresses for each node and one for iDrac/IPMI. For example, if you have a four-node cluster, reserve twelve IP addresses. Four IP addresses are required for node IP addresses, four for Virtual IP address (VIP) addresses, and four for iDrac/IPMI.
If you configure VIPs on any other VLAN, then the number of VIPs must match the number of nodes in the cluster. This option provides full load balancing and also high availability.
iDrac and IPMI IP address
- The iDrac interface is used for remote management. This interface is essential for pulling hardware diagnostic logs when needed for hardware related support issues. The iDrac interface IP address is then also used for IPMI
- The Intelligent Platform Management Interface provides console-level access over IP and is also used for node-to-node cluster level hardware monitoring.
- Reserve IP addresses for the iDrac or IPMI interface or iDrac interface. Most hardware platforms
require one IP address for each node.Draft comment: Shalini.Sood2@ibm.comDo we need this note?
Note: Intelligent Platform Management Interface (IPMI) is a standard for controlling intelligent devices that monitor a system. It is helpful for dynamic discovery of sensors in the system, the ability to monitor the sensors, and be informed when the sensor's values change or go outside certain boundaries. Integrated Dell Remote Access Controller (iDRAC) is Dell's implementation of IIPMI. - The 1 GbE interface on each node is used for the iDrac network traffic and requires a dedicated IP address.
- All nodes on the cluster require iDrac/IPMI interface IP addresses that are in the same subnet. All the IPMI or iDrac IP addresses must be in the same subnet, either in the IPMI/iDrac subnet or the cluster subnet.
- Each node's IPMI interface must be able to reach the IP addresses of all other IPMI interfaces on the cluster.
Node IP address
- Node IP address is used for internal cluster traffic and also for cluster management connectivity.
- Each node has two 10 GbE ports that are bonded for failover and do not require any special switch port configuration.
- By default, the two 10 GbE ports are configured for Adaptive Load Balancing (ALB).
- One node IP address is assigned to the pair of bonded physical 10 GbE ports on each node.
- Other than the gateway, each node must also be able to reach Domain Name System (DNS) and a non-Windows Network Time Protocol (NTP) server.
- All node IPs must be in the same subnet as the VIPs and must be able to reach the backup sources.
- Enable multicast traffic on your network (required for the IBM® Storage Defender Data Protect auto-discovery function).
- Nodes can also communicate over tagged VLAN that is non-native VLAN.
- Open firewall ports to allow the cluster to transmit and receive data. For more information, see
Manage Firewall Ports in the IBM Storage Defender Data Protect User
Guide in the IBM Storage Defender Data Protect reference information.Note: To access the Data Protect reference information, you must authenticate by using IBMid credentials that are associated with your IBM Storage Defender account.
Virtual IP address
- Each node is also assigned a VIP on the bonded 10 GbE interface.
- The VIP is used for backup traffic, file services (SMB, NFS, S3), and for accessing the IBM Storage Defender Data Protect management interface.
- All VIPs on a cluster are registered with DNS under a single hostname.
- VLAN tagging is supported on VIPs.
DNS setup
- All VIPs on a cluster are registered with DNS under a single hostname.
- DNS entries are not required for node IPs.
- To effectively share traffic between all the nodes on the cluster, the DNS server circulates the hostname resolution among all the VIPs of the cluster. For more information, see Configuring a DNS server.
If you configure cluster VIPs in primary interface instead of VIPs in VLAN, then you can assign VIPs less than the number of nodes. In such a setup, only a few nodes with VIPs can receive VIP traffic. This option still provides high availability for the configured VIPs. James:I’m not exactly sure what this means. I’ve sent a Q to our QA to see how many VIP’s are actually needed.