Weird MAC/Vlan related issues with multiple uplinks

[commy]

Hi All,

We have a WS-26-500-DC running 1.5.0rc4 and have many vlans passing through 2 separate uplinks.

The uplinks are connected to the same switch on the other side passing the same vlans (yes together you will cause a loop, we are aware) so what we do when we want to switch over from one link to the other we just enable the Ethernet port on the netonix of the first link and disable it on the second by checking/unchecking the appropriate checkboxes.
Upon switching manually no traffic passes through the link as if it appears that the MAC table is not updated at all. Waiting for 10 minutes makes no difference and only a reboot of the netonix fixes this issue. After rebooting the netonix correctly starts passing traffic as it should through the link that was brought up and had issues. The ethernet is correctly brought up and shows that its running so this is not a cable/interface issue.

This method of switching used to work fine on a Mikrotik CRS and a Cisco 2960G with the same type of simple configuration.
Can anyone help as to why the netonix has this issue? We are aware that this is not an efficient way to switch uplinks but we have it for special purposes and need it like this.

Thank you.

[Stephen]

If you could, please post screenshots of all the tabs on your switch. Specifically the Status, Ports, and VLAN tabs, but the more you post the more it will help.

[commy]

Attached requested tabs.

LAG/STP/QOS/ERPS are all disabled or n/a


Ports 1 and 15 are the ones that are identical behind radios that end on the same remote switch. In the photos you will see that the active port is 15 and port 1 is disabled.
If we enable port 1 and disable port 15, then save - we have to reboot in order for the switch to pass traffic appropriately.

Thanks

[mike99]

Config on the remote site switch ?

[commy]

Its a MikroTik CRS125, nothing fancy, both ports are on the same bridge.

[mike99]

Both ports are in trunk mode so the vlan tab is not use. In trunk mode, it's the allowed vlan list that is use. Click on trunk check bot at the top vlan to see allowed vlan for the port.

[commy]

The trunk allowed list have the same vlans as the ones that are checked - this goes for all ports. So essentially the ones that have trunk port checked in their lists have the same vlans that are marked as "T".

[commy]

Anyone some other idea ?

Thanks

[mike99]

More informations would help even if those seem simple. Could you give us Mikrotik interface config including VLAN ID with wireing schema + the Netonix trunk config ?

[bayamon]

Try making a small change in VLAN config (e.g. make some unrelated port T on some unrelated VLAN), save it, and change it back.

I've seen many time this fixes a correctly configured netonix that refuses to pass VLAN traffic.