Help with connecting M6348 to a Cisco Network

Question

Hi Can anyone help me with best practice in connecting R710's via a stacked M6348 to a cisco network. Our configuration is as follows M1000e 2 x 1GB Passthrough in the A Fabric 2 x M6348's cabled in a stack in the B Fabric 6 x R710's with Quad-port Nics in the B Fabric (3 are general purpose servers, 3 are Hyper-V Servers in a cluster)   Cisco Network Mostly Cisco 3560 switches 20 VLANS (1 per building) configured, mostly for traffic isolation (VLAN 300 - 320)   What (I think) we want to do Connect the M6348's to the core network in a LAG (4 x 1 GB ports. Can this be split 2 ports from each of the stacked M6348's?) Have a straight mapping for R710 1 - 3 to VLAN 301 Configure R710 4 - 6 to access all VLAN's so the Hyper-V guests can be placed on the correct VLAN Use Fabric A for connection to the SAN Network   Out network management is done by a third party (we are unable to change this) and they claim they are unable to connect Cisco/Dell! We are server guys, not network guys, so something idiot proof (especially for the Cisco to give to our management company) would be appreciated. We have tried configuring the switches in Simple mode, using a Deployment guide from Dell, but all we seem to get is the default VLAN to every port on switch 1 of the stack, but no other VLANs and no ports active on switch 2 in the stack. I have probably missed lost of info out and haven't been very clear, so if any more info is needed, please get back to me. Thanks in advance for the help Rich

ridgewoodschool · Answer

That's brilliant, thank you. Do I need to manually configure the VLAN's on the 6348 stack, or will they appear automatically from the connection to the Cisco core once the LAG is running?

ridgewoodschool · Answer

Thanks for all your help so far. Finally seem to be making progress.

Just a few more questions.

When configuring the port-channel, I got en error with

channel-group 1 mode auto

% Invalid input detected at '^' marker

The options I seem to have are active or on. The switch is running v 5.1.1.7.

Our network provider has told me that the default VLAN should be VLAN 301. What do I need to do for this?

I was planning on using server 2012 to team 3 of the 4 nics on the switch and configure the team as switch independent/hyper-v port. Is this a sensible configuration, or would you recommend switch dependent? None of the individual VM's should need more than a single NIC's bandwidth.

ridgewoodschool · Answer

OK, I think im getting there. There just a few more questions.

Do I need to use a different channel-group for each server team/LAG group?

This is what I have so far:-

!Configure Vlan

vlan database
vlan 300-320

!Commands to name VLAN's ommitted

!Configure generl purpose servers
interface Gi1/0/1
switchport mode access
switchport access vlan 301

interface Gi1/0/2
switchport mode access
switchport access vlan 301

interface Gi1/0/3
switchport mode access
switchport access vlan 301

!Configure Hyper-V ports (repeat for other 2 servers with different channel group number?)
interface Gi1/0/10
channel-group 2 mode active <---- Ive used channel-group 2, as I am already using 1 for the link to the core. Is this correct?

interface Gi1/0/26
channel-group 2 mode active

interface Gi2/0/10
channel-group 2 mode active

interface Gi2/0/26
channel-group 2 mode active

!Configure LAG to Core network
interface Gi1/0/33
channel-group 1 mode active

interface Gi1/0/34
channel-group 1 mode active

interface Gi2/0/33
channel-group 1 mode active

interface Gi2/0/34
channel-group 1 mode active

!Configure Hyper-V Channel group VLAN's
interface channel-group 2 <--- This give me an error. Should it be interface port-channel 2?
switchport mode trunk <--- Is this needed?
switchport trunk allowed vlan add 300-320
switchport trunk native vlan 301

!Configure Link to core switch VLAN's
interface channel-group 1
switchport mode trunk
switchport trunk allowed vlan add 300-320
switchport trunk native vlan 301

ridgewoodschool · Answer

OK. All configured on the Dell side (I think) but the servers are not able to communicate with the core network.

I am hoping that this is because the network team havnt configured the core switch LAG correctly, and im not getting all the VLAN's, but is there any way to check from the Dell side that this is the case?

ridgewoodschool · Answer

Config below

Ive removed sensitive data, and unnecessary lines. Ive added *** where I thought it best to leave something in that I may have introduced an error.

Rich

!Current Configuration:

!System Description "PowerConnect M6348, 5.1.1.7, VxWorks 6.6"

!System Software Version 5.1.1.7

!System Operational Mode "Normal"

!

configure

vlan 300-305,307-315,317-319,324,400

exit

vlan 300

name "***"

exit

vlan 301

name "***"

exit

vlan 302

name "***"

exit

vlan 303

name "***"

exit

vlan 304

name "***"

exit

vlan 305

name "***"

exit

vlan 307

name "***"

exit

vlan 308

name "***"

exit

vlan 309

name "***"

exit

vlan 310

name "***"

exit

vlan 311

name "***"

exit

vlan 312

name "***"

exit

vlan 313

name "***"

exit

vlan 314

name "***"

exit

vlan 315

name "***"

exit

vlan 317

name "***"

exit

vlan 318

name "***"

exit

vlan 319

name "***"

exit

vlan 324

name "***"

exit

vlan 400

name "ClusterTraffic"

exit

slot 1/0 1 ! PowerConnect M6348

slot 2/0 1 ! PowerConnect M6348

! stack

member 1 1 ! PCM6348

member 2 1 ! PCM6348

exit

interface vlan 1

exit

interface vlan 300

exit

interface vlan 400

exit

username "***" password *** privilege 15 encrypted

!Removed Shutdown interfaces

!

interface Gi1/0/10

channel-group 2 mode active

exit

!

interface Gi1/0/26

channel-group 2 mode active

exit

!

interface Gi1/0/33

channel-group 1 mode active

exit

!

interface Gi1/0/34

channel-group 1 mode active

exit

!

interface Gi2/0/10

channel-group 2 mode active

exit

!

interface Gi2/0/26

switchport access vlan 400

exit

!

interface Gi2/0/33

channel-group 1 mode active

exit

!

interface Gi2/0/34

channel-group 1 mode active

exit

!

interface port-channel 1

description "Core Network LAG"

switchport mode trunk

switchport trunk native vlan 301

exit

!

interface port-channel 2

description "Hyper-V01"

switchport mode trunk

switchport trunk native vlan 301

exit

ridgewoodschool · Answer

Sorry, the copy/paste was a little aggressive in removing data. Ive added again below

Port-channel 2 should have 3 ports in it. These ports are all from the same server. On the server (2012R2) they are configured as a teach (switch dependent/LACP)

interface port-channel 1
description "Core Network LAG"
switchport mode trunk
switchport trunk native vlan 301
switchport trunk allowed vlan 300-305,307-315,317-319,324
exit

interface port-channel 2
description "RWS-HV-01"
switchport mode trunk
switchport trunk native vlan 301
switchport trunk allowed vlan 300-305,307-315,317-319,324
exit

ridgewoodschool · Answer

Ive managed to get the config that the support team have put into the cisco core switch. Its shown below.

Is there any way to test the configuration from the switch to make sure everything is running correctly. It doesn't appear to be working as I aren't able to get any communication from the server to any of the network. This could be that the server is not configured correctly, so is there any testing I can do on the powerconnect switch to verify the configuration so I know where to look for errors?

!
interface Port-channel1
description SERVER
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
!
interface GigabitEthernet0/1
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/2
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/3
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/4
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/5
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/6
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/7
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!
interface GigabitEthernet0/8
description SERVER LACP group 1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 300-324
switchport mode trunk
channel-protocol lacp
channel-group 1 mode active
!

ridgewoodschool · Answer

Done some additional testing today, and still not much further.

From the 6348, verified that I could ping devices on each VLAN. All OK. But not sure this test is much use, as all it means is that there is connectivity to the core switch. Also, there is a VLAN 1, which I guess is coming from the M1000e chassis, and I worry that this is the link that is being used for this traffic, so maybe nothing is actually going through the trunk to the core.

Other tests I have done.

Server 10 is a test server. I have taken all cards out of the team.

On Gi1/0/10, I set the port to general, pvid 301 (default VLAN)

I manually gave the server an IP address on the 301 vlan. (10.49.56.202)

I tried to ping from the 6348 to this server, and got no response.

I couldnt ping anything (other than itself) on the same vlan.

Im at a loss how to make any futher progress

ridgewoodschool · Answer

OK, some more testing. Blade 10 Configured Gi1/0/10 as access Assigned to vlan 400 Gave an IP address of 192.168.200.10   Blade 11 Configured Gi1/0/11 as access Assigned to vlan 400 Gave an IP address of 192.168.200.11   Tried pinging between the blades - worked fine. Removed Blade 11 from vlan, pinged to verify loss of connectivity.   Changed VLAN from 400 to 301 for Gi1/0/10 and Gi1/0/11 Changed IP addressing for Blade 10 and 11 to 10.49.56.210 and 10.49.56.211 respectively. Pinged from 10.49.56.210 to 10.49.56.211 - Successful Pinged from 10.49.56.211 to 10.49.56.210 - Successful Pinged from 10.49.56.210 to default router (Cisco Layer 3 switch) on 10.49.56.3 - No response Pinged from 10.49.56.211 to default router (Cisco Layer 3 switch) on 10.49.56.3 - No response IP addressing for Vlan 301 is 10.49.56.0/24

ridgewoodschool · Answer

I can arrange for a port to be configured on the cisco as a general connection. The SLA for changes is 5 days, so may take a while.

IP addressing for VLAN 301 is correct. Network address is 10.49.56.0/24

IP address for default router on the VLAN 301 is 10.49.56.2

DHCP allocates addresses in the range 10.49.56.10 to 10.49.56.100

Blade 10 is currently configured as follows:-

IP Address 10.49.56.210/24

Default gateway 10.49.56.2 (This is the IP address of the layer 3 switch (Cisco core)

ridgewoodschool · Answer

OK So configured the port-channel in general as suggested. I am now able to ping the core network from blade 10.

ridgewoodschool · Answer

But is it OK to leave it like this? Surely it will need to be configured as a trunk?

Networking General

Was this post helpful?