vlan routing woes 7024

Hi,

VLAN 1 is routable on that firmware version, unlike some of our other switches VLAN 1 is not the management network since there is an OOB port. So VLAN 1 is treated the same as any other VLAN. Switchport general combines an access port and trunk port and can handle both untagged and tagged traffic. If all your traffic is tagged on the trunk ports there is not an need to use a general port. Page 796 ftp://ftp.dell.com/Manuals/all-products/esuprt_ser_stor_net/esuprt_powerconnect/powerconnect-7048_Reference%20Guide_en-us.pdf

Do the trunk ports have all of the VLANs added to them? Unlike some other vendors our trunk ports deny all VLANs by default and each VLAN has to be added.

cheers for the prompt feedback. I'd seen reference to vlan1 not being routable on other Dell series switches.

Both 7024 switches have matching VLAN configurations (diff ip addresses in the same subnets obviously). Config snippets below for a little clarity.

Vlan 111 has its own dedicated trunk. Connectivity between devices across sites over this vlan is fine.

interface Te1/1/1  

switchport mode trunk  

switchport trunk native vlan 111  

switchport trunk allowed vlan 111

interface Te1/1/2  

switchport access vlan 111

Wishing to pass all other traffic over the other trunk. No effect by setting vlan 1 as the native vlan on this trunk.

interface Te1/2/1  

switchport mode trunk  

switchport trunk allowed vlan 1-10,999

If you run a show interfaces status does it show Te1/1/2/1 as up? Setting native vlan 1 should allow tagged and untagged vlan 1 traffic across the trunk. IS routing is enabled globally with the ip routing command and the routing command on that trunk port?

On that interface

Console #routing

Hi,

ip routing is enabled globally.

Link is up

#show interfaces status Te1/2/1
Port       Name                       Duplex  Speed    Neg   Link   Flow Control                                                                    State  Status
---------  -------------------------  ------  -------  ----  ------ ------------
Te1/2/1    OS1 trunk               Full    10000    Off   Up     Active

Can you clarify the routing command I should be using for that trunk ?  

this is where the confusion starts and one of the reasons for the firmware upgrade (from 5.1.0.1 to 5.1.3.7).

sw1(config-if-Te1/2/1)#routing
                                 ^
% Invalid input detected at '^' marker.

Similarly, I've seen reference to vlan routing or vlan routing vlanid indexid within the vlan database content - in various example configs for other Dell switches as well as the 7000 Series Systems CLI Reference Guide from Oct 2011 (A03). Strangley it doesn't appear in the Mar 2013 (A04) CLI ref guide. It looks like it should work but gives an error.

sw1(config)#vlan database

sw1(config-vlan)#vlan routing ?

                     Press enter to execute the command.

sw1(config-vlan)#vlan routing

VLAN range - separate non-consecutive IDs with ',' and no spaces.  Use '-' for range.  Specify 1024 VLANs in the range <1-4093>.

sw1(config-vlan)#vlan routing 999
                                       ^
% Invalid input detected at '^' marker.

I've just cross checked the two ref guides - the routing command isn't listed in Mar 2013 (A04).

Any clarification appreciated.

It does look like it was removed in the later firmware versions, and it automatically applied when a VLAN is created. If you just put one VLAN on the trunk does it work? If you add these VLANs to the working trunk?

I'd initially configured this trunk purely for vlan 1 using switchport trunk native vlan or allowed vlan commands. With no success, further testing then done with the additional vlan 999 to no avail. Both switches at either end of the link configured with the same vlans, unique ips for the relevant subnets and matching port configurations for these trunk links.

interface Te1/2/1 

switchport mode trunk 

switchport trunk allowed vlan 1-10,999

It seems to be configured correctly, which is why it is odd that it isn’t allowing pings or connections, if you assign ip addresses to the ports can they ping each other?

i need to know how to conncect by cli to this router.

spammer ?

this is not a router

RTFM for cli access.

Belated update to this query - temporarily at least, I've gone down a different path, and got no further with this.  However, I noted entries in the logs indicating the my problem above is actually a spanning tree issue.

I don't understand why STP is getting involved, trunk ports are not sharing vlans at all - why don't I have the required isolation with the below config (snippet):-

interface Te1/1/1  

switchport mode trunk  

switchport trunk native vlan 111  

switchport trunk allowed vlan 111

interface Te1/1/2  

switchport access vlan 111

interface Te1/2/1  

switchport mode trunk  

switchport trunk allowed vlan 1-10,999

No other ports configured for vlan111.

Some other ports (for systems mgnt) using switchport access vlan 1.

Do the three ports above plug into the the same device? Do you have a brief topology diagram you can show us?

no topology diagram available.

Matching Dell 7024 switch at the other end with a near identical setup

interface Te1/1/1 &  interface Te1/2/1

fiber-based links back to primary site

interface Te1/1/2  

connected to NAS device at each site

vlan 111 connected to only those two NAS units and the uplink between these two sites - nothing else.

on a different port, a single connection to vlan 1 on the core switch.

OK, so if I understand correctly the two connecting switches have two connections between them?

If they are setup like this, there will be a loop in the network and stp will step in. I would suggest making the two connections into a LAG and set the LAG as a Trunk connection.

Hi,

not quite there. Re-reading the thread, the description given wasn't necessarily self-explanatory. Vlan111 was working and initially thought is was a vlan1 routing issue.

Te1/2/1is a second trunk not Access as per your diagram.

Te1/1/1 & Te1/2/1 are two separate trunks connecting two switches each over 10G fiber and servicing totally different VLANs.  I wanted the vlan111 (iSCSI or NFS NAS comms) traffic isolated/separate from all other traffic, rather than using a LAG over those 2 10G links for all VLANs.  Alas, with the current config STP steps in. I'm still curious to know where I've gone wrong.

If I had the correct LX SFP+ transceivers I would have connected the two NAS systems direct.