Port Config when migrating from Cisco to HP Comware

[DennisL1]

Hello,

I'm having a problem migrating a specific port config from a Cisco Switch to an HP 5700 series Comware 7 based switch.

The attached device is an appliance based on Syneto Storage OS with a specific NIC configuration:

- The NIC has a management IP, which cannot be assigned a VLAN tag.

- Additional VLANs can however be created, thus creating a trunk where packaged from a VLAN are tagged with the VLAN ID, but traffic from management IP remains untagged.

The port config on the Cisco is:

interface GigabitEthernet0/47
  switchport access vlan 3
 switchport trunk encapsulation dot1q
 switchport trunk native vlan 131
 switchport trunk allowed vlan 131,141
 switchport mode trunk
 switchport nonegotiate
 spanning-tree portfast

The "switchport trunk allowed vlan 131,141" allow traffic from vlan 131 and 141, while the "switchport trunk native vlan 131" basically tells the switch that all packages coming in the port should be treated as being vlan 131 (wich is our management VLAN)

On the HP I have tried this:

interface Ten-GigabitEthernet1/0/17
 port link-type trunk
 undo port trunk permit vlan 1
 port trunk permit vlan 131 141
 port trunk pvid vlan 131
 stp edged-port
 #

So basically I also allow vlan 131 and 141, and from an HP document I found that "port trunk pvid vlan 131" sets the native VLAN to 131, and that the native VLAN in untagged

However this is not working. If I put the LAN cable from the Cisco to the HP, we can no longer ping the management IP

Both Switches have uplinks to the same network, other devices connected to these switches are accesible just fine, the issue is specifically related to the port config of this device.

I also tried to make this a hybrid port, with this config:

interface Ten-GigabitEthernet1/0/17
  port link-type hybrid
 undo port hybrid vlan 1
 port hybrid vlan 141 tagged
 port hybrid vlan 131 untagged
 port hybrid pvid vlan 131
 stp edged-port
#

But that didn't solve the problem either.

Does anyone have an idea on how I should configure the HP port so that it matches exaclty the Cisco config, because the device has been working for >2 year while connected to the Cisco...

Thank you,

Dennis

[sdide]

Do you have link-up?

I mean, you are going from a gigabit interface to a ten-gigabit interface. So the interface needs to accept the 1-gig transceiver.

What does the log say, when you plug the transceiver in, and after that, what does

]display transceiver interface te1/0/17

output,

and maybe better also

]probe
-probe] display transceiver information interface te1/0/17

Regards

[DennisL1]

Hello,

Yes, the link is active. The server actually has a 10 Gig interface, which works "only" on 1 Gb on the Cisco switch, and on 10 Gb with the new HP switch.

I believe the problem is related to the tagged/untagged vlans and a difference between the cisco and HP in the way thay handle tagged/untagged traffic...

[sdide]

Hi,

That is strange.

What software are you running on the 5700?

interface Ten-GigabitEthernet1/0/17
 port link-type trunk
 undo port trunk permit vlan 1
 port trunk permit vlan 131 141
 port trunk pvid vlan 131
 stp edged-port

should work exactly as you want.

Since HP has no Dynamic trunking protocol, that could have been an issue, but you have "switchport nonegotiate" on the Cisco device, so - its not that. 
Try clearing the arp entry on the router. Maybe the Synogo box is not chatty by default, and doesn't send out gratuitous ARP on linkup.

Regards.

[DennisL1]

Hello,

thanks for the response. It's running HPE Comware Software, Version 7.1.045, Release 2422P01

That's not the latest release I know.

There is no router involved. The switches are interconnected using a bridge-aggregation link. Perhaps I can clear the arp on the old switch, haven't tried that yet.

I'm also a bit dependent on our system engineer who manages the server, to move the patching back and forth between the old and new switch ...

[Vince-Whirlwind]

The odd thing is in your Cisco port config, you have "switchport access vlan 3". What does this do? Seeing as it is configured as a trunk port, you shouldn't have any "access" config on there.

Rather than repatching the server, I would test this with a laptop.

Give the laptop an IP address in the VLAN131 subnet, then patch it to the HP switchport. Your HP config *looks* good, but it could be that your network doesn't work exactly the way you think it does.

Once your laptop in VLAN131 can ping other VLAN131 addresses, you will know your untagged VLAN is working.

[sdide]

Hi,

the "switchport access vlan 3" statement shouldn't have any significance as long as the "switchport mode trunk" is set. (you'd need "switchport mode access" for that)

I dunno why Cisco IOS does not clear that line on the interface when you change mode. But - It shouldn't do anything.

Regards

[DennisL1]

Thanks all for the suggestions and feedback.

I've just put a laptop on the switchport and gave it an IP address in de VLAN131 range. It was working perfectly fine, so that indeed proves packets which are untagged (the laptop offcourse has no vlan ID set) are send to VLAN131 by the switch. So far so good.

I still have to try to ARP suggestion but the system admin is currently out of the office, and I'm not going to repatch that system without him present ;)

PS: by default all our switchports are configured as "switchport access vlan3" VLAN3 is actually not used and us just a way to block traffic to the port (on top of a mandatory shutdown on all unused ports)

For some reason on Cisco switches, when you change the operating mode van switchport mode access to switchport mode trunk, the access vlan remains in the configuration. I leave it there in the event a ports is switched back to access mode...

[DennisL1 · Accepted solution]

Status Update:

It works!!

What did I we change? --> Nothing !

So, We basically patched the Syneto server back to the HP switch, after which we lost ping response to the management IP on VLAN 131

Then I cleared the ARP tables on all involved switches, but to no avail.

So, while I was discussing options with the system engineer, on how we would proceed and what tests we could perform to resolve the issue, after 10 or 15 minutes we started getting ping responses without any of us making modifications.

I can only conclude that there is some ARP cache or something like that on the syneto system which is not performing as expected and some automatic periodic arp clearing on the machine finally resolved our issue.

Thanks all for confirming I was getting my port config properly and for pointing me in the right direction.

[merlinbrent]

glad its working!

Just FYI (and as you seem to be moving from Cisco to HP, this wont matter) but

running:

no switchport access vlan 3

will clear the line of config, so you won't see it. This could help to avoid confusion when looking at the port configs.

Thanks