NSX-T 2.5.1 Segmet Uninitialized

[KaBalint]

Hello Everyone!

I am facing an issue, i can create everything on the Advanced Networking and Security tab, everything works perfectly, but when i changed to the Simplyfied UI and I've created a Segment it tells me: Uninitialized.

I've created a separated Edge, and Edge Cluster for the T0 GW, also created a separated Transport Zone too.

The NSX-T version is 2.5.1 all the three managers work well. I have tested the Advanced N&S with a working PKS deployment.There are 4 transport nodes, in a Shared Edge and Compute deployment.

Is there anybody who had seen this problem?

Kind Regards,

KaBalint

[daphnissov]

If you don't create a T0/T1 in the policy-based UI then you won't be able to utilize the logical segment.

[KaBalint]

I have created a T0/T1 GW too on the Policy Based UI. Those gw-s are up with green dot, but when i connect them to the segment it show up an exclamation mark with realization error, doesnt matter it is a downlink, or an external link.

[daphnissov]

You may need to check your transport nodes and ensure there are no issues. Sounds like there's a problem either at the edge or one or more of the nodes.

[KaBalint]

I think all of the nodes work well, because on the advanced ui, i can create everything, and all of the configurations work well.

Anyway thanks for your advice, these are all good ideas!

[mauricioamorim]

Are you sure you have host TNs with the necessary TZ? When you say you created a new TZ you would have to associate it with your host TNs. Realization errors usually occur if NSX manager doesn't have the necessary resource to provision the segment.

[KaBalint · Accepted solution]

Im sorry, so the TZ means a vlan-tz, which needed on the edge to the uplink. Because on the uplink i wanted to set up a vlan tag, and on the port group which is connected to the edge was a trunk pg.

Anyway the shut down (all the three NSX-TManagers at the same time) and a turn on solved the problem.

Thanks for your help!

[daphnissov]

So then the cause was like I suspected: Something in one or more of the hosts was in an inconsistent state. A reboot can often fix these things like you've found. The fact that you shut them all down simultaneously shouldn't have made a difference.

[KaBalint]

When i restarted the nsx managers one-by-one did not help, thus this was my last chance to shutdown all of the managers, and it worked.

[daphnissov]

Oh so you shut down all of the manager appliances. Ok, that must have been something else then. Like a database replication/consistency issue.

[KaBalint]

Yep, im sorry, i was not clear enough im my reply. I am going to edit it if it possible.