HPE Storage Users Group

PaulC wrote:I've attached a file that shows the connectivity between the 3Par and my Itaniums through the switches. as you can see everything not colored is plugged in but not seen be the 3Par.

any thoughts as to why or if there is further info needed please let me know.

Best practices for Port Persistence do not appear to be followed. Typically you would have all EVEN numbered 3PAR ports on 1 switch, and all ODD numbered ports on the other. This implies your zoning could be incorrect as well. Was this system relocated and reconnected somewhere new?

Your current SW1:
20:01:00:02:ac:01:a6:e1
21:02:00:02:ac:01:a6:e1
22:01:00:02:ac:01:a6:e1
23:02:00:02:ac:01:a6:e1

Your current SW2:
20:02:00:02:ac:01:a6:e1
21:01:00:02:ac:01:a6:e1
22:02:00:02:ac:01:a6:e1
23:01:00:02:ac:01:a6:e1

Recommended SW1 (and verify update zoning):
20:01:00:02:ac:01:a6:e1
21:01:00:02:ac:01:a6:e1
22:01:00:02:ac:01:a6:e1
23:01:00:02:ac:01:a6:e1

Recommended SW2: (and verify update zoning):
20:02:00:02:ac:01:a6:e1
21:02:00:02:ac:01:a6:e1
22:02:00:02:ac:01:a6:e1
23:02:00:02:ac:01:a6:e1

When you move ports around from one switch to another, or from one VSAN to another, you will need to change/configure your zoning.

'show zone'

PaulC wrote:I debated adding to this or creating a new post if I should do that please let me know.
it is a continuation.

I've attached a file that shows the connectivity between the 3Par and my Itaniums through the switches. as you can see everything not colored is plugged in but not seen be the 3Par.

any thoughts as to why or if there is further info needed please let me know.

the switches are Cisco 9148s

thanks

What Richard said.

And.

20:xx:xx... WWNs on fc1/1, fc1/2, fc1/3 and fc1/4 are port WWNs of virtual connect. These are not initators in a FC fabric and should (and will) not be visible on the 3PAR.

Same goes for "sup-fc0" as that is the switch itself.

That leaves 2x 3PAR ports... (3:0:2 and 0:0:1) which shouldn't be seen by the 3PAR anyways... which makes me wonder where you see the other 6 host ports... anyways ...

and it leaves FACS1 and FACS3BL870 on SW1....

I would do as Richard said, start with checking zoning on that switch.

As long as the ports are up and you get those WWNs out from the name server on the switch, then it's a zoning issue (assuming the 3PAR is healthy and hasn't failed over persistent ports and thrown them into the other fabric.. but then you wouldn't have the WWNs from the 3PAR on that switch).

hmmm the cabling I did all the top ports on the 3Par and connected to the top switch sw1 and the bottom to sw2, seemed to make the most sense. here are the 2 zones, and I threw everything at it to make sure it could see it all.

also even if I have extra WWN's shouldn't it still see it and connect?


NHSANSW1# show zone
zone name facst vsan 2
pwwn 50:01:43:80:01:69:74:84
pwwn 20:01:00:02:ac:01:a6:e1
pwwn 21:01:00:02:ac:01:a6:e1
pwwn 22:01:00:02:ac:01:a6:e1
pwwn 23:01:00:02:ac:01:a6:e1
pwwn 50:01:43:80:16:78:61:7c
pwwn 50:01:43:80:16:78:61:7e
pwwn 50:01:43:80:24:29:37:88
pwwn 50:01:43:80:24:29:37:8a
pwwn 20:11:00:11:0a:00:4a:6f
pwwn 20:12:00:11:0a:00:4a:6f
pwwn 20:13:00:11:0a:00:4a:6f
pwwn 20:14:00:11:0a:00:4a:6f
pwwn 50:01:43:80:33:14:f5:4c
pwwn 20:12:00:11:0a:00:04:05
pwwn 20:13:00:11:0a:00:04:05
pwwn 20:14:00:11:0a:00:04:05
pwwn 20:15:00:11:0a:00:04:05
pwwn 20:16:00:11:0a:00:04:05
pwwn 20:17:00:11:0a:00:04:05
pwwn 20:00:00:11:0a:00:04:05
pwwn 20:15:00:11:0a:00:4a:6f
pwwn 20:16:00:11:0a:00:4a:6f
pwwn 20:17:00:11:0a:00:4a:6f
pwwn 20:00:00:11:0a:00:4a:6f
pwwn 50:01:43:80:18:73:cc:30
pwwn 50:01:43:80:18:73:cc:32
pwwn 50:01:43:80:21:dc:ec:7c
pwwn 50:01:43:80:21:dc:ec:7e
pwwn 50:01:43:80:09:ab:9e:e6
pwwn 50:01:43:80:09:ab:9e:e4
pwwn 50:06:0b:00:00:c2:62:00
pwwn 50:06:0b:00:00:c2:62:02
pwwn 50:01:43:80:33:14:f5:4e

zone name facs1 vsan 2
pwwn 20:15:00:11:0a:00:04:05
pwwn 20:16:00:11:0a:00:04:05
pwwn 20:17:00:11:0a:00:04:05
pwwn 20:00:00:11:0a:00:04:05
pwwn 20:11:00:11:0a:00:4a:6f
pwwn 20:12:00:11:0a:00:4a:6f
pwwn 20:13:00:11:0a:00:4a:6f
pwwn 20:14:00:11:0a:00:4a:6f
pwwn 20:15:00:11:0a:00:4a:6f
pwwn 20:16:00:11:0a:00:4a:6f
pwwn 20:17:00:11:0a:00:4a:6f
pwwn 20:00:00:11:0a:00:4a:6f
pwwn 50:06:0b:00:00:c2:62:00
pwwn 50:06:0b:00:00:c2:62:02
pwwn 50:06:0b:00:00:c2:62:04
pwwn 50:06:0b:00:00:c2:62:06
NHSANSW1#

my OS is telling me there is an issue with the following connections, the last 3 in yellow didn't show up till a restart of the switch.

SYS$PGQDRIVER - QLogic ISP23xx - FGG0 - SPDT 91643440 - STDTs
-------------------------------------------------------------
| STDT Cred Port FP NP Appr | Act Cmd Cnf Cls | QF Seq
STDT FC-LA Port Name LID PortID | Stat I/Os I/Os I/Os I/Os I/Os | Sus Sus Pnd Pau | Seen TMOs
-------- ----- ------------------- --- ------ + ---- ---- ---- ---- ---- ---- + --- --- --- --- + ---- ----
91643CC0 00001 2301.0002.AC01.A6E1 082 070200 | 0001 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 00BF
9164B4C0 00002 2101.0002.AC01.A6E1 083 070300 | 0001 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0042
94878C80 00005 2002.0002.AC01.A6E1 081 070800 | 0001 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000
9487F200 00006 2102.0002.AC01.A6E1 FFF 000000 | 01A4 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000
94874A80 00007 2202.0002.AC01.A6E1 085 070900 | 0001 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000
94876080 00008 2001.0002.AC01.A6E1 FFF 000000 | 01A4 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000
94876C80 00009 2201.0002.AC01.A6E1 FFF 000000 | 01A4 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000
94874C80 0000A 2302.0002.AC01.A6E1 FFF 000000 | 01A4 0000 0000 0000 0000 0000 | 000 000 000 000 | 0000 0000

Who taught you to zone?

Read this and try again

https://en.m.wikipedia.org/wiki/Fibre_Channel_zoning

no one! well maybe about 18 years ago someone may have but nothing since

ok so the only things on these switches are my 3 nodes and the 3Par and they are all together, I'm ok with that there is nothing else on it.

thanks! I'll take a look

PaulC wrote:no one! well maybe about 18 years ago someone may have but nothing since

ok so the only things on these switches are my 3 nodes and the 3Par and they are all together, I'm ok with that there is nothing else on it.

thanks! I'll take a look

If you put everything into one zone you might as well disable zoning all together.

If not, one host and one storage system (or one target and one initiator) per zone and multiple zones in a zone set. Aliases on WWNs to make them human readable.

Cisco has multiple alias types. There is a friendly name in the DCNM gui, that is not on the switches, there is a fcalias that is tied to the VSAN, and there can also be a device-alias that spans VSANs. Admins who use DCNM can achieve human readable zoning via the GUI but the config on the switch still be wwn based zoning.

Cisco does have an Smart Zone method, but that is not what was documented above. It is similar to Brocades it's similar to Brocade's Target driven peer zoning, in a sense that the zone has ports marked as targets, and ports marked as initiators, and the switch auto-resolves this to apply a single target-single initiator relationship, not allowing initiators to see other.

PaulC - I recommend getting some professional services to help you sort this out, and get you on track. HPE, and most VARs, have consultants that can help with both the 3PAR and the switches. Fibre Channel switching is robust and low latency, but it's also complex. Mistakes can easily lead to outages when you don't 2 redundant fabrics as a starting point, and not following best practices.

I believe your fabric needs some redesign too, top of mind is that you are using VSAN2 on both switches. This puts you at high risk of an accidental zone-merge if anyone accidentally runs a cable from switch 1 to switch 2, and will cause some headaches cleaning that up.

https://www.cisco.com/c/en/us/support/d ... tches.html