They are at different layers really.

 

Broadcast domains describe network L2 port connectivity. Failover groups describe rules for selecting port to relocate LIF.

 

You should create separate broadcast domain for ports that are connected to a separate L2 network (VLAN). You can subdivide broadcast domain in multiple failover groups if you have different requirements for relocating ports (let’s say 1GbE and 10GbE ports - you probably want LIF on 10GbE port to failover to another 10GbE port, not to 1GbE).

 

Failover groups are part of broadcast domains. Failover groups cannot span multiple broadcast domains. Same applies to subnets - there could be multiple subnets defined inside of the same broadcast domain (not a good practice, but it happens) but subnet cannot span multiple broadcast domains.

 

So if “exclude” in your question refers to pure cluster administrator desire - create additional failover group. If your port will be connected to a different network - it is almost certainly is connected to different L2 domain and so you need to create another broadcast domain to start with.

 

Of course you are free to ignore all of the above J and manage networking in old way by using single default broadcast domain and failover groups inside of it. I do not see anything in documentation that would prohibit it (and if you update from previous releases you end up with this configuration). It just that new broadcast domain abstraction makes it easier to organize your ports to match your networking topology and to speak with your networking team using the same language J

 

 

From: toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of Momonth
Sent: Sunday, May 24, 2015 2:45 AM
To: Basil
Cc: toasters@teaparty.net
Subject: Re: FAS8040: SP and e0M configuration

 

That all turned out to be not so obvious, I ended up re-configuring the cluster from scratch, that way the SP and e0M worked straight away after 'cluster setup' completion.

 

So the idea of having "ipspaces" / "broadcast domains" is to let OnTAP to configure "failover groups" automatically for you.

 

Now, let's say, I need to exclude a physical port from the existing "Default" failover group, what do I need to employ:

 

failover-groups modify

or 

broadcast-domain remove-ports

 

?

 

On Sat, May 23, 2015 at 4:22 PM, Basil <basilberntsen@gmail.com> wrote:

You'll want to configure broadcast domains- each of them will come with a gateway and a pre-built failover group.

 

On Sat, May 23, 2015 at 9:48 AM, Momonth <momonth@gmail.com> wrote:

Good point, I'm actually running:

na102::> version
NetApp Release 8.3P1: Tue Apr 07 16:05:35 PDT 2015

I'll dig into respective 8.3 docs, thanks for the hint!

Cheers,


On Sat, May 23, 2015 at 3:42 PM, Basil <basilberntsen@gmail.com> wrote:
> Assuming you're not using 8.3 which has new logic in place for networking,
> you'd want to create a routing group:
> https://library.netapp.com/ecmdocs/ECMP1196907/html/GUID-BA68EFA4-CFEA-468B-B3F2-FBCD39B63E6B.html
>
> On Sat, May 23, 2015 at 9:07 AM, Momonth <momonth@gmail.com> wrote:
>>
>> Hi All,
>>
>> I never used e0M on FAS32xx as it was limited to 100Mb, but decided to
>> give a try on FAS8040.
>>
>> So, I've configured SP via DHCP (10.193.12.0/24), it's up and running
>> and I can login onto it:
>>
>> na102::> sp show -node na102labnode-1a -instance
>>
>>                                Node: na102labnode-1a
>>                      Type of Device: SP
>>                              Status: online
>>               Is Network Configured: true
>>                   Public IP Address: 10.193.12.31
>>                         MAC Address: 00:a0:98:68:00:18
>>                    Firmware Version: 3.1P1
>>                         Part Number: unknown
>>                       Serial Number: unknown
>>                     Device Revision: Not Applicable
>>      Is Firmware Autoupdate Enabled: true
>>
>>
>> However having assigned an IP to e0M (from cluster management
>> interface) doesn't make the e0M "ping-able". Here is my config (the
>> same 10.193.12.0/24 is used):
>>
>> na102::> network interface show -vserver na102 -lif
>> na102labnode-1a_mgmt1 -instance
>>
>>                     Vserver Name: na102
>>           Logical Interface Name: na102labnode-1a_mgmt1
>>                             Role: node-mgmt
>>                    Data Protocol: -
>>                        Home Node: na102labnode-1a
>>                        Home Port: e0M
>>                     Current Node: na102labnode-1a
>>                     Current Port: e0M
>>               Operational Status: up
>>                  Extended Status: -
>>                          Is Home: true
>>                  Network Address: 10.193.12.44
>>                          Netmask: 255.255.255.0
>>              Bits in the Netmask: 24
>>                  IPv4 Link Local: -
>>                      Subnet Name: -
>>            Administrative Status: up
>>                  Failover Policy: local-only
>>                  Firewall Policy: mgmt
>>                      Auto Revert: true
>>    Fully Qualified DNS Zone Name: none
>>          DNS Query Listen Enable: false
>>              Failover Group Name: Default
>>                         FCP WWPN: -
>>                   Address family: ipv4
>>                          Comment: -
>>                   IPspace of LIF: Default
>>
>> It feels like e0M needs default gateway? How do I set it up then?
>>
>> Or Am I missing the whole idea here?
>>
>> Thanks,
>> Vladimir
>> _______________________________________________
>> Toasters mailing list
>> Toasters@teaparty.net
>> http://www.teaparty.net/mailman/listinfo/toasters
>
>