10 Gig interfaces limited to 1 gig vlan.

[Kerry Kriegel]

Hello,

I have the following setup in my data center for connectivity from multiple cores to multiple edges (each edge connects to multiple providers).

Edge-1  Ten 2/1 <-----------> Core-1  Ten 9/1 (VLAN 961)

Edge-2  Ten 2/1 <-----------> Core-1  Ten 9/2  (VLAN 961)

Edge-1  Ten 2/2 <-----------> Core-2  Ten 9/1 (VLAN 958)

Edge-2  Ten 2/2 <-----------> Core-2  Ten 9/2  (VLAN 958)

Edge-1  Ten 2/3 <-----------> Core-3  Ten 2/1 (VLAN 959)

Edge-2  Ten 2/3 <-----------> Core-3  Ten 2/2  (VLAN 959)

Edge-1  Ten 2/4 <-----------> Core-4  Ten 2/1 (VLAN 960)

Edge-2  Ten 2/4 <-----------> Core-4  Ten 2/2  (VLAN 960)

These are 6500’s and their 10-G connections to each of the 2 edge 6500’s is a direct fiber connection.
Please note that the interfaces are not trunking, and the edges do not have any spanning-tree instances running.
The edges run BGP with the providers and the cores are running EIGRP within the datacenter.

This is a snippet of the "show interface" for all of the VLAN 958 connections ..

Edge-1

#sh int te 2/2

TenGigabitEthernet2/2 is up, line protocol is up (connected)

  Hardware is C6k 10000Mb 802.3, address is d0d0.fd0e.fc40 (bia d0d0.fd0e.fc40)

  Description: Connected to DC2-Core-A TenGigabit 9/1 via FO1-B-2.

  Internet address is 66.185.29.194/28

  MTU 1500 bytes, BW 10000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 12/255, rxload 13/255

  Encapsulation ARPA, loopback not set

  Keepalive set (10 sec)

  Full-duplex, 10Gb/s, media type is 10Gbase-SR

Edge-2

#sh int te 2/2

TenGigabitEthernet2/2 is up, line protocol is up (connected)

  Hardware is C6k 10000Mb 802.3, address is 0011.5d96.cc00 (bia 0011.5d96.cc00)

  Description: Connected to DC2-Core-A TenGigabit 9/2.

  Internet address is 66.185.29.195/28

  MTU 1500 bytes, BW 10000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 3/255, rxload 10/255

  Encapsulation ARPA, loopback not set

  Keepalive set (10 sec)

  Full-duplex, 10Gb/s, media type is 10Gbase-SR

 Core-1

#sh int te 9/1

TenGigabitEthernet9/1 is up, line protocol is up (connected)

  Hardware is C6k 10000Mb 802.3, address is 001a.a10e.d844 (bia 001a.a10e.d844)

  Description: Connected to DC1-Edge-A TenGigabit 2/2 via FO1-B-2.

  MTU 1500 bytes, BW 10000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 13/255, rxload 12/255

  Encapsulation ARPA, loopback not set

  Keepalive set (10 sec)

  Full-duplex, 10Gb/s, media type is 10Gbase-SR

#sh int te 9/2

TenGigabitEthernet9/2 is up, line protocol is up (connected)

  Hardware is C6k 10000Mb 802.3, address is 001a.a10e.d845 (bia 001a.a10e.d845)

  Description: Connected to DC2-Edge-A TenGigabit 2/2.

  MTU 1500 bytes, BW 10000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 11/255, rxload 3/255

  Encapsulation ARPA, loopback not set

  Keepalive set (10 sec)

  Full-duplex, 10Gb/s, media type is 10Gbase-SR

#sh int vlan 958

Vlan958 is up, line protocol is up

  Hardware is EtherSVI, address is 00d0.02da.3400 (bia 00d0.02da.3400)

  Description: Connected to the EIGRP Edge-cloud.

  Internet address is 66.185.29.193/28

  MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 249/255, rxload 148/255

  Encapsulation ARPA, loopback not set

  Keepalive not supported

  ARP type: ARPA, ARP Timeout 04:00:00

 ..

..

All of the VLAN's are showing the same bandwidth of 1 gig.  Why is the vlan interface speed is only 1 gig?

Each of the VLAN's is configured with a /28 so I can make each individual interface a layer 3 if necessary.

I can upload a more detailed pdf drawing if it is needed.

[Joseph W. Doherty · Accepted solution]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

My guess, it probably is a default, because what should a SVI interface's bandwidth be?  I mean, would you want it to be a constantly changing aggregate of all the interfaces that are "up" in the VLAN?  Or, how should a trunk interface be counted, as it's shared with other VLANs?

Assuming the SVI accepts a bandwidth statement, configure it to what is useful/correct for you.

[Kerry Kriegel]

Joseph,

It appears you are correct.  The "default" bandwidth is one gig.  I have a maintenance window scheduled for next Friday night and will take the opportunity to execute the 'bandwidth' command on the SVI's.

Thank you

[jackvestal]

Hey Kerry,

I'm seeing the same thing here and wondering what became of you running your "bandwidth" command.  Make any difference one way or the other in regards to SNMP reporting?

[Madhukrishnan Gopinathan Nair]

Vlan interfaces BW is logical and it will not drop packets even if it goes over 1 Gig.

So i believe there is nothing to worry here.

Thanks

Madhu

[Joseph W. Doherty]

Right, packets shouldn't be dropped, but SMNP tools could show utilization over 100%.

[stolar]

Sorry if I am making old topic once more alive, but I've been searching long time and I didn't find any plausible answer. Anyone can explain me what maximal throughput has an Interface Vlan on a Catalyst L3 Switch? I know BW is a logical parameter used i.e. in EIGRP, I have some SVI's which are running over 1Gb/s and no packets are dropped. The question is how to get information if such Int Vlan goes over his capabilities? How to monitor it over SNMP while all of NMS which I know grabs BW parameter which is only 1Gb/s and creating false alarms...

[Joseph W. Doherty]

To my knowledge, a SVI, itself, shouldn't have any physical limit.

Something like a bandwidth statement might have a logical configuration limit.

For monitoring purposes, either set the SVI to reflect the aggregate of all the physical ports it controls (on the SVI's physical device) or to the maximum logical configuration limit.  (The latter would be likely easier to manage.  However, if the latter is insufficiently large, either stop monitoring the SVI or request an enhancement from Cisco.)

[bbb bbb]

well, I tried to create a new VLAN interface. and succeeded in changing the bandwidth