Move PS-100 Equallogic from one subnet to another

Typically reconfiguring the IP of an array is done through the serial interface, most data centers use a Digi PortServer to do this remotely.  The reason for the serial connection is that Telnet/SSH/HTTP sessions could drop when the eth interfaces are disconnected and reconnected.

With that said, I would configure the management network first.  Testing that you can ping each member eth2 and management group IP from the work station to ensure you have connectivity.

Then you can go ahead and configure the new IP’s, however you need to do one member eth at a time, so do member 1(eth1) to the new subnet, then do member2 (eth1).  Then change the group IP, then configure member 1 (eth0) and member2 (eht0).  The reason for this is that each member must have at least one active Ethernet interface on the Array that is in the same subnet as the group address at all times.

You can give support a call, they have a document on how to do this in greater detail.

-joe

Thanks Joe,

I've updated the management eth2 and changed the ip address and default gateway on both nodes. Now when I connect via HTTP I get the logon box and after entring the username / password get "connection to the server could not be established" - regardless of opening as application (java) or in the browser. ???

I can't connect via the other eth(1) and (0) addreses either (still on 192.168.1.x)

I have connected to the array via serial cable and all members are up (eth0 / eth1 / eth2 (which shows as management-enabled)) so not sure why I am unable to connect anymore. The new switch I have plugged one of the boards into is 100Mps (unlike current 1Gbp) but figure thats no problem

Thanks Joe,

after that then I should be able to connect via http to the Managment-netowrk IP address? (not the individual IP addresses on the eth2 controllers for each array.

So until this is configured I won't be able to connect to the eth0 and eth1 ports and Group IP address (still 192.168.1.x). It seems regardless what I have done with the management netowrk, I should still be able to connect to the Group IP? - or not?....

Once you configure the highest port with a IP on a different subnet, the other SCSI ports will no longer allow you to connect, so this is normal.

Check the following:

Via the serial connection on the active controller, configure the management Group IP address.

> grpparams management-network ipaddress 10.x.x.x

(change this to your actual IP, this IP must be unique and must be a different IP then you will use for the eth2 ports on either member)

This will show the address once configured:

> grpparams management-network show

Now ensure the eth interface is setup for management only:

> member select membername eth select 2 ipaddress 10.x.x.x netmask 255.255.x.0

> member select membername eth sel 2 up

> member select membername eth sel 2 mgmt-only enable

(change to your actual IP address, the IP must be unique and different then the IP used for eth2 on the other member)

Do the eth config for both members (change the member name)

The following command shows the details for the management-only interface (check both members):

> member select membername eth select 2 show

-joe

Once the management network is enabled, you use the "Management Group IP" to connect to the group (this is the IP you setup with the grpparams management-network ipaddress 10.x.x.x command).  At this point, you use the Group management IP to manage the group.  The highest eth setup for management are typically not used, however there are some situations were support will what you to login to a specific member, at that point you would use the management port on the member.

The Group IP (different then the management group IP) is then only used by the host server iSCSI initiator to discover targets. Once the host initiator has discovered the target and attempts to log in, the connection will be redirected to a physical eth on the iSCSI subnet.

-joe

do both eth2 ports need to be live/up on the switch in order to connect via managemtn-group ip? or can you access the group with one down?

IGEUSANGR01> grpparams management-network show

Ip Address

----------------------------------------------------------------

172.16.4.81

member01> member select member01 eth sel 2 show

_______________________________ Eth Information _______________________________

Name: eth2                             Status: up

Changed: Thu Feb 16 11:26:59 2012      Type: ethernet-csmacd

DesiredStatus: up                      Mtu: 1500

Speed: 100 Mbps                        HardwareAddress: 00:09:8A:01:48:35

IPAddress: 172.16.4.77                 NetMask: 255.255.255.0

IPv6Address:                           Description:

SupportsManagement: yes                ManagementStatus: enabled

_______________________________________________________________________________

member01> member select member02 eth sel 2 show

_______________________________ Eth Information _______________________________

Name: eth2                             Status: up

Changed: Mon Jan 23 09:43:08 2012      Type: ethernet-csmacd

DesiredStatus: up                      Mtu: 9000

Speed: 1 Gbps                          HardwareAddress: 00:09:8A:01:A7:15

IPAddress: 172.16.4.78                 NetMask: 255.255.255.0

IPv6Address:                           Description:

SupportsManagement: yes                ManagementStatus: enabled

_______________________________________________________________________________

member01>

I can't ping anything from my desktop? - but I can ping when on the switch. Need to checking my routing

ok all set up and pinging the managment ports (eth2 on both arrays) correctly and also the management group IP, but upon loggin in getting the same error after entering the passowrd "connection to the server could not be established"

I set the eth2 members back to the 192.168.x.x subnet and disable dedicated management, I could reconnect correctly, it seems when making it dedicated on another subnet I get log into via http.

for the connection to server not being established, try both telnet and ssh, one should work.

-joe

All management interfaces should be connected (the active and standby controllers eth ports on both arrays).  The reason for this is if, for some reason, there is a group lead change and the active eth interface is on longer plugged in... well you get the idea, you would lose management connection to the group via IP (however serial is still always available).

-joe

hmm, but I really need to be able to connect via http.

If I change the group IP (not mgmt) and default gateway and can't change it back, I'm in trouble. I can't find any CLI commands on changing the default gateway on the iSCSI interfaces...

Both Arrays are on F/W 5.0.2, would an upgrade to the latest help?"

Actually, I think this is it:

def-mgmt-gateway ipaddress|none

Configures or removes the default gateway for the group (IPv4 only).

If you can connect via telnet/ssh to the group via the management IP, then something else is blocking http.

(I forgot to add the management default gateway command as part of the management network configuration, you do need this as well).

Otherwise (regarding http). check that you have the following ports open on any router between your desktop and array (also, ensure all three management IP's are not blocked):

TCP 80 HTTP To group management IP address

TCP 80 HTTP To each eth management IP address

TCP 3002 GUI communication To group management IP address

TCP 3003 GUI communication (encrypted) To group management IP address

Note there are other port that would need to be open if using SNMP, etc.  These are listed in a KB on the support site.

A simple test is to connect a pc/laptop directly to the switch on the management vlan, and see if you can connect there, is so then you know it's a routing issue.

-joe

I also managed to connect  to the new group management ip via SSH. But I really need http