Hi, I am having some problems getting VLANing to work correctly. I was hoping someone would be able to point me in the right direction. I have an RV220W switch which has 2 VLANs on it VLAN1 with a subnet of 192.168.1.x and VLAN 104 with a subnet of 192.168.104.x. Port 4 of the goes to port 20 of the SG300 Switch. Port 20 on the SG300 switch has VLAN 1 as untagged and VLAN104 as tagged. It also has port 15 and 16 configured the same way.
In my ESXi host NIC2 is in port 16 and NIC3 is in port 15. I have created a Standard vSwitch1 and created two port groups. Group 192.168.104.x has a vlan id of 104 and group vmware with no vlan id. I would assume that when I put a virtual machine in the port group vmware I would get an ip address of 192.168.1.x which I do.
When I put a virtual machine in port group 192.168.104.x I would assume that I should get an ip address of 192.168.104.x but I do not, I get a loop back address.
Any ideas what I am doing wrong? I am assuming that by creating the port group 192.168.104.x with a VLAN id of 104 that this is the untagged native vlan for this switch, which would then send the traffic, back through port 15 or 16 as tagged VLAN104. Which would then send that traffic out port 20 and into port 4 of the RV220W which would act as the DHCP server for this replying as 192.168.104.1.
Just when I think I have a grip on VLANs I am wrong.
Thanks
I would recommend this:
Configure an access port on the SG300, with no trunking
Configure it to VLAN 104
Place a desktop on that VLAN
Configure your desktop and VM with a static IP
Test if they have connectivity
This will rule out VMware configuration, and confirm that your trunk to the SG300 is operational.
As a furthur FYI, I tried putting a static IP address of 192.168.101.100 on the virtual machine. I am able to ping its own ip address but I am not able to ping the default gateway of 192.168.1.1
After speaking with Cisco Tech support, they have verified that my Router and Switch Configs are correct.
The means it comes down to one question. Does a standard virtual Switch support tagging and untagging of 802.1q?
TdisalvoOrinoco wrote:
I am assuming that by creating the port group 192.168.104.x with a VLAN id of 104 that this is the untagged native vlan for this switch
Actually it would mean traffic would land on the physical switch tagged with VLAN 104.
Standard switches absolutely do support dot1q tagging.
Josh,
Thanks so much for repling. Any idea what I have configured wrong? My though is that the traffic should go from the virtual machien nic to the virtual swtich port group, there get tagged with the vlan id. then get that passed out the physical nics on the esxi host out to the sg300 switch port 15 or 16, It shoudl arrive there tagged as vlan 104 then be sent across all ports that allow vlan 104tagged traffic which would in turn send it out port 20 of the sg300 and into port 4 of the rv220. There the rv220 should see that it is vlan 104 and know that it is the default gateway for this traffic and route it accordingly, be it replying to the dhcp request, or just routing traffic.
I cant even put on static addresses and have them route to the default gateway:-(
Hi,
If I've understood everything correctly it all looks right.
I've never runs VLANs on entry level hardware like this with much success, I don't suppose you have any other router or switch to try playing with?
Sadly no, this is a lab for a small consulting compnay so $ are at a premium.
This swtich was almost 400.00, I know that a Cisco 2960 is the prefered switch of choice but I would need a definative answer stating that it was the switch that was blocking this from working. The switch is new so I can still return it, but Cisco is saying that the switch config is good.
Cisco needs a Switch rental program, like home depot has for tools.
Thanks for your input. You don't happen to know any other good sites for VLAN info for an issue like this, by any chance do you.
I have also opened up a case with VMware, it is one of my two cases that I get. Hopefully they will be able to resolve this and I will post the resolution up here.
I would recommend this:
Configure an access port on the SG300, with no trunking
Configure it to VLAN 104
Place a desktop on that VLAN
Configure your desktop and VM with a static IP
Test if they have connectivity
This will rule out VMware configuration, and confirm that your trunk to the SG300 is operational.
Josh,
Thanks so much for your help on this issue. That definatly proved that the issue was on the firewall/router. Once I had that isolated I was able to go back to cisco and show them that DHCP was not even getting out to the physical pc configured on VLAN 104.
An interesting side note. VMware was clearly tagging the packets correctly. I found that if I put a static ip address on a vm in the 104 vlan and one on the pc with the port in the 104 vlan, from vmware, I was able to ping the physical machines ip addres, however the machine was not able to ping the virtual machines ip address. So my understanding is that the virtual switch was tagging the traffic correctly on its way out, this is why it was able to ping out.
But the phiscal switch was not tagging correctly otherwise it would have been two way traffic, since they are onthe same physical switch.
I now have to wait for a beta firmware upgrade to see if they can resolve the intra vlan routing issue. I will try to update this post with the conclusion of my experince. Hopeully someone else will be able to find it when they have a simmilar issue.
As a final note on this issue. Cisco did end up replacint the RV220W router, and now everything is working great.
Thanks to everyone for all of their help on this issue.