First crack at FC using MDS9100

[Steven Williams]

I have entered the world of storage networking this past week and need to get MDS9100's up and running for a Lab EMC VNX. I am very versatile in LAN/WAN and also Nexus, just not storage tasks within the Nexus platform. The MDS seems to be NX-OSish and I am familiar with most commands on it. What I have learned so far watching some videos is:

Zoning is like ACLs in the R&S world. Zones allow certain ESXi hosts or servers to see certain storage. So I have about 15 luns all part of the same storage pool. So will I have 15 WWPNs? One for each lun?

vSANs are like vlans, I just dont understand when to have multiple vSANs within an environment?

What are some basics I need to know to get storage and servers talking? WWPNs from the servers and storage, I need to know Port type, I assume all port types will be N ports. 

Do MDSs work in like a HA pair? So I connect each MDS to each other and trunk them? I understand there isnt any load balancing in FC like there is in IP when using something like vPC?

[Steven Williams]

So one pair of MDS9148s run both VNX5300(LAB) and VNX5400 (PROD) also all Lab ESX hosts and PROD Hosts are connected to the same MDS. So all Prod hosts can see PROD VNX and LAB VNX and also all Lab hosts can see PROD VNX and LAB VNX. Its a mess. Everything is vsan 1 on both MDS's.

So now PCI auditors come in to play and say "Hello, your Prod and Lab equipment cannot share the same fabric!" so we bought new MDS9148s and I need to split the environments.

So I need to setup the new MDS's for Lab Hosts and Lab VNX5300.

But right now the lab hosts are only using 4 SP's total on the lab VNX5300 because the Prod hosts are using the rest, so when we remove the Prod hosts, we have "extra" SP's that I think we should use since we can.

[dynamoxxx]

makes sense.  So before you disconnect LAB VNX from prod MDS, what are you going to do with all the storage on the LAB VNX that is being used by prod ESX ?

[Steven Williams]

There wasnt much bleed over, even though all hosts could see all storage they kept it somewhat neat. But anything thats left will be storage vmotioned.

[dynamoxxx]

got it.  So yeah, when you move LAB VNX to LAB MDS, i would defintly try to take advantage of the extra ports. If you look at the example i provided above you will see how i am alternativing my ESX servers between 2 x 4 port sets on VNX.

[Steven Williams]

Right, I agree but so I need to keep the config the same on the new MDS, like the zoning needs to be the same for the cluster to work correctly? 

[dynamoxxx]

how you move into this new lab environment depends if you want to take full downtime or do a rolling migration/vMotion/svMotion. 

[Steven Williams]

I have a full 12 hours on monday. I assume i need to zone the same way is it currently is not to have to mess with the EMC portion of it. Otherwise I am sure I need to make changes to the zone masking.

[dynamoxxx]

correct, if you simply want to swing VNX ports and ESX ports to the new MDS then you need to zone it the same way it's zoned in production. If you do decide to "clean up" a bit and start using more ports than that will be more involved as most likely you will need to change the zoning and initiator registration on VNX. It's your call.  You can migrate as it is right now to satisfy your PCI auditors and then slowly work on your lab environment (put node into maintenance mode, re-zone, re-register on VNX, move VMs over ..rinse and repeat)

[Steven Williams]

Does EMC put their config guides on their website?

[dynamoxxx]

what specifically are you loooking for ?

[Steven Williams]

How to use more SP ports. How to re zone on the EMC side.

[dynamoxxx]

You will find "Host Connectivity Guides" very helpful, i don't think they address the zoning. Login to support.emc.com and look around in the documentation section. Zoning is simple, you want to use as many ports as you think your application will take advanage of. On VNX you want your host to be zoned to at least 4 ports.

[Steven Williams]

nevermind, now that I drew it on paper is see what your doing. You are alternating SP ports per MDS. 

[dynamoxxx]

if this were my lab this is how i would zone to take advantage of all ports on VNX . Notice how i am alternating VNX ports between ESX servers.

MDS 1

 zone name esx1-h0-vnx5600ndb2 vsan 10

    pwwn 50:06:01:60:36:b0:12:13 [vnx5600-spa0]  target

    pwwn 50:06:01:64:36:b0:12:13 [vnx5600-spa4]  target

    pwwn 50:06:01:69:36:b0:12:13 [vnx5600-spb1]  target

    pwwn 50:06:01:6d:36:b0:12:13 [vnx5600-spb5]  target

    pwwn 20:00:00:25:b5:01:a1:2f [esx1-h0]  init

  zone name esx2-h0-vnx5600 vsan 10

    pwwn 50:06:01:62:36:b0:12:13 [vnx5600-spa2]  target

    pwwn 50:06:01:66:36:b0:12:13 [vnx5600-spa6]  target

    pwwn 50:06:01:6b:36:b0:12:13 [vnx5600-spb3]  target

    pwwn 50:06:01:6f:36:b0:12:13 [vnx5600-spb7]  target

    pwwn 20:00:00:25:b5:01:a2:9f [esx2-h0]  init

 zone name esx3-h0-vnx5600ndb2 vsan 10

    pwwn 50:06:01:60:36:b0:12:13 [vnx5600-spa0]  target

    pwwn 50:06:01:64:36:b0:12:13 [vnx5600-spa4]  target

    pwwn 50:06:01:69:36:b0:12:13 [vnx5600-spb1]  target

    pwwn 50:06:01:6d:36:b0:12:13 [vnx5600-spb5]  target

    pwwn 20:00:00:25:b5:01:a3:2f [esx3-h0]  init

  zone name esx4-h0-vnx5600 vsan 10

    pwwn 50:06:01:62:36:b0:12:13 [vnx5600-spa2]  target

    pwwn 50:06:01:66:36:b0:12:13 [vnx5600-spa6]  target

    pwwn 50:06:01:6b:36:b0:12:13 [vnx5600-spb3]  target

    pwwn 50:06:01:6f:36:b0:12:13 [vnx5600-spb7]  target

    pwwn 20:00:00:25:b5:01:a4:9f [esx4-h0]  init

MDS 2

  zone name esx1-h1-vnx5600 vsan 20

    pwwn 50:06:01:63:36:b0:12:13 [vnx5600-spa3]  target

    pwwn 50:06:01:67:36:b0:12:13 [vnx5600-spa7]  target

    pwwn 50:06:01:6a:36:b0:12:13 [vnx5600-spb2]  target

    pwwn 50:06:01:6e:36:b0:12:13 [vnx5600-spb6]  target

    pwwn 20:00:00:25:b5:01:b0:2e [esx1-h1]  init

  zone name esx2-h1-vnx5600 vsan 20

    pwwn 50:06:01:61:36:b0:12:13 [vnx5600-spa1]  target

    pwwn 50:06:01:65:36:b0:12:13 [vnx5600-spa5]  target

    pwwn 50:06:01:68:36:b0:12:13 [vnx5600-spb0]  target

    pwwn 50:06:01:6c:36:b0:12:13 [vnx5600-spb4]  target

    pwwn 20:00:00:25:b5:01:b1:5f [esx2-h1]  init

  zone name esx3-h1-vnx5600 vsan 20

    pwwn 50:06:01:63:36:b0:12:13 [vnx5600-spa3]  target

    pwwn 50:06:01:67:36:b0:12:13 [vnx5600-spa7]  target

    pwwn 50:06:01:6a:36:b0:12:13 [vnx5600-spb2]  target

    pwwn 50:06:01:6e:36:b0:12:13 [vnx5600-spb6]  target

    pwwn 20:00:00:25:b5:01:b2:2e [esx3-h1]  init

  zone name esx4-h1-vnx5600 vsan 20

    pwwn 50:06:01:61:36:b0:12:13 [vnx5600-spa1]  target

    pwwn 50:06:01:65:36:b0:12:13 [vnx5600-spa5]  target

    pwwn 50:06:01:68:36:b0:12:13 [vnx5600-spb0]  target

    pwwn 50:06:01:6c:36:b0:12:13 [vnx5600-spb4]  target

    pwwn 20:00:00:25:b5:01:b3:5f [esx4-h1]  init

[Steven Williams]

This whole thing to me is confusing. I am looking at the initiator section of the VNX, and my Labesxi host has 1 HBA (2Ports) but yes it is connected to 4 SP ports? How is this possible?