Configuring management ports for Data Migrator to Cloud

Data Migrator Administration Guide for Hitachi NAS Platform

Version
15.1.x
Audience
anonymous
Part Number
MK-92HNAS005-33
Best practice is to use the aggregate ports instead of the management ports. Aggregate ports transfer data with a higher outgoing performance rate. However, if you plan to use the management ports for cloud targets, you must define the following:
  • DNS
  • Networking
Note: The Virtual Secure Servers feature is not compatible with Data Migrator to Cloud.
CAUTION:
Before using Data Migrator to Cloud, you must ensure that the managements ports are properly configured.

Data Migrator to Cloud connects to the cloud through one of the management ports on the system and not through the aggregate ports that are used for NFS or CIFS access. Because of this, you must configure the eth0 or eth1 network interface card with an IP address, netmask, and gateway that is routable to the cloud target.

  • If the cloud target is Hitachi Content Platform, a route must exist to the namespace that will be configured as part of the cloud account.
  • If the cloud target is Amazon S3, a route must exist to https://s3.amazonaws.com/.
In a clustered environment, usually one network interface card is configured for the cluster interconnect (eth1) and the other (eth0) is not configured. In general, the cluster interconnect is a private network among the cluster nodes and it is expected that the cloud traffic will be sent through eth0. For example, eth1 can be used for the cluster interconnect and eth0 for the cluster node IP, as illustrated in the graphic:


Hitachi Global Solutions and Services recommends the following network configuration for cloud traffic:

Single node network configuration:


Clustered network configuration


The first figure displays a single node network configuration and the second a clustered network configuration. Each node requires two IP addresses:
  • One cluster node IP address on eth0 and eth1.
  • One Admin EVS IP address on eth1.
    • This IP will failover between the nodes as needed so it is not important which node it is configured on.
The following tables show examples of an IP address configuration:
Single Node
Interface IP
SMU eth0 172.10.10.11
Cluster Node IP eth0 172.10.10.13
Cluster Node IP eth1 172.10.10.14
Admin EVS IP eth0 172.10.10.15
Cluster
Interface IP
SMU eth0 172.10.10.11
Cluster Node IP eth0 172.10.10.13
Cluster Node IP eth1 172.10.10.14
Admin EVS IP eth0 172.10.10.15
2 Cluster Node IP eth0 172.10.10.16
2 Cluster Node IP eth1 172.10.10.17
Note: The SMU does not require redundant network connectivity.

All IPs configured on eth1 and eth0 of the HNAS node(s) should be on the same subnet.

To configure a cluster node IP addresses on a management port, issue the following commands on the CLI:

ipaddr [-r] [-p <port>] [-I <addr> -m <mask>] [-c | --confirm]

For example, to add the address 192.168.53.116 to eth1 with netmask 255.255.248.0, issue the following CLI command:

ipaddr -p eth1 -I 192.168.43.116 -m 255.255.248.0

This requires DNS in order to resolve the cloud target. To configure DNS, the CLI dnsserver command can be used:

dnsserver add <server1> <server2>

For example the following command assigns DNS servers 192.168.45.10 and 192.168.45.11

dnsserver add 192.168.45.10 192.168.45.11

Make sure to configure a default route for the eth0 and eth1 interfaces. This can be done with theCLI route command:

route add gateway -g <addr>

For example the following command assigns default gateway 192.168.43.1.

route add gateway -g 192.168.43.1

Note: When the cloud target is HCP, Global Solutions and Services recommends the HCP and HNAS eth0 and eth1 interfaces reside on the same subnet.

These commands must be run once before configuring the cloud objects such as account and destination, and these settings are persistent across any reboot of the system. To modify or delete these settings, consult the man pages for the ipaddr and route commands.

Note: Using the ipaddr and route commands will not modify other addresses and routes already configured on the management interfaces.
Tip: To verify your routing and DNS configuration, the CLI ping and host commands can used.