AFF

ned help on zoning

ozonne94
4,701 Views

Dear community,

 

Is anyone can help on this please

I have already made zoning but on entry level storage and no NPIV. I am trying to understand the Netapp logic.

I have 2 AFFA200 in cluster mode  each storage is located on different room separate by 5 km.

storage1 present 2 lun 10 and 11

storage 2 present 2 lun 12 and 13

each esxi has 2 HBA and should see all lun with 4 path /per HBA is it correct ?

Can you give me an exemple on how to zone correctly ?

Here is the configuration.

Thank you !

 

Netapp-zoning (1).jpg

 

1 ACCEPTED SOLUTION

TMACMD
4,441 Views

I will go back to my first post: NPIV is a must.

 

In Clustered ONTAP (not 7-mode), ONTAP NEVER-EVER uses the hardware assigned address (50:xx...) and ALWAYS uses the Logical assigned address (20:xx...)

Or as the docs put it:

 

You must use the WWPNs of the target LIFs, not the WWPNs of the physical FC ports on the storage nodes. The LIF WWPNs are all in the range 2x:xx:00:a0:98:xx:xx:xx, where x is any hexadecimal digit. The physical port WWPNs are all in the range 50:0a:09:8x:xx:xx:xx:xx.

 

Read more here:

https://docs.netapp.com/ontap-9/topic/com.netapp.doc.exp-fc-cpg/GUID-9D2AAC92-A160-4C01-B381-C6892342E071.html

 

View solution in original post

5 REPLIES 5

TMACMD
4,660 Views

NPIV is a must.  ontap will only work with fiber channel LUNs when using NPIV

 

 

ozonne94
4,610 Views

Hello,

I know thanks but it doesn't answer to my question 🙂

AlexDawson
4,527 Views

General principle is "single initiator zoning" - ie, you have a zone for each initiator.

 

If you're not doing boot from SAN, this isn't totally needed for ESX - you could just create one big zone, and the default config should work. But I'd recommend doing it anyway..

 

So, you've created the aliases for each of the targets and initiators, now create a zone on each fabric which contains the initiator and both of the targets it should see. Then apply the zoneset, and rescan in ESX and you should see it.

ozonne94
4,497 Views

Thanks for your answer

I think my zoning is correct now but

when I initiate a ping to ESXi hba 1 wwwn here is what I have

 

 

CLU01::> vserver fcp ping-initiator show -vserver CLU01_FC 51:40:2e:c0:15:57:30:30
Node Logical Ping Extended
Vserver WWPN Name Interface Status Status
--------- ----------------- ---------- --------- ----------- -----------------
CLU01_FC
51:40:2e:c0:15:57:30:30
FAS01
FAS01_fc_lif-0e
reachable wwpn-logged-in
51:40:2e:c0:15:57:30:30
STOFAS01
STOFAS01_fc_lif-0f
not-zoned -
51:40:2e:c0:15:57:30:30
FAS02
FAS02_fc_lif-0e
reachable wwpn-logged-in
51:40:2e:c0:15:57:30:30
FAS02
FAS02_fc_lif-0f
not-zoned -
4 entries were displayed.

CLU01::>

 

the strange thing is that in the zone all these target are present and also the initiator (ESXi hba1)

example of my zone

 

ESX01_S2P1_svm_STOCLU_FAS01_FAS02_FAS03_FAS04
50:0a:09:81:80:f4:23:2b
20:08:d0:39:ea:29:92:fd
50:0a:09:81:80:f4:1b:c3
20:01:d0:39:ea:29:e3:bd
50:0a:09:81:80:c4:1b:00
20:03:d0:39:ea:29:e3:bd
50:0a:09:81:80:64:18:e6
20:07:d0:39:ea:29:92:fd
51:40:2e:c0:15:57:30:30

 

TMACMD
4,442 Views

I will go back to my first post: NPIV is a must.

 

In Clustered ONTAP (not 7-mode), ONTAP NEVER-EVER uses the hardware assigned address (50:xx...) and ALWAYS uses the Logical assigned address (20:xx...)

Or as the docs put it:

 

You must use the WWPNs of the target LIFs, not the WWPNs of the physical FC ports on the storage nodes. The LIF WWPNs are all in the range 2x:xx:00:a0:98:xx:xx:xx, where x is any hexadecimal digit. The physical port WWPNs are all in the range 50:0a:09:8x:xx:xx:xx:xx.

 

Read more here:

https://docs.netapp.com/ontap-9/topic/com.netapp.doc.exp-fc-cpg/GUID-9D2AAC92-A160-4C01-B381-C6892342E071.html

 

Public