LDR issue? (Setup screenshots added)

[tanagi]

Hello,

In a new install, I am unable to ping across ldr.

Setup is as follows,

2 vDS (compute and mgmt)

3 clusters (mgmt, compute a and b)

Single controller (NSX ver 6.0.4)

VXLAN: all clusters in VLAN 100, mtu 1600, single VTEP per cluster

Single transport zone with 3 clusters

3 logical switches: comp-a (5000), comp-b (5001) and mgmt (5002); unicast control plane

ldr with 3 interfaces corresponding to each switch.

IP addresses of ldr interfaces do not show up under "Manage/Interfaces" (web client)

NSX edge is not listed when trying to add it from switches (like VMs)

VMs are able to ping within clusters but not across ldr.

Any guidance is appreciated.

[tanagi]

Made following change with bit of progress... changed LDR to ESG. Am able to ping local and remote router interfaces, but not the VMs' interface IPs (in different subnets across ESG).

Here are screen shots of set up,

NSX/Installation/Management:

NSX/Installation/Host Management:

VXLAN Transport:

Transport Zone:

Logical Switches:

ESG:

ESG does not show up to be added to respective switch,

ESG interfaces added as follows,

What am I missing?

[SpasKaloferov]

Hi ,

i would suggest first to sure this is not the cause of your issue. This turned out to be the cause for many LDR communication issues recently

NSX VXLAN Enable Agent fails on ESXi hosts with error “Cannot complete the operation.”

http://kaloferov.com/blog/nsx-vxlan-enable-agent-fails-on-esxi-hosts-with-error-cannot-complete-the-...

BR,
Spas Kaloferov

[tanagi]

Thanks SK. Will check and get back.

[mrlesmithjr]

@tanagi - Were you able to sort this out? I myself have been going through a very similar issue since around 12.20.2014 and still seeing the same issues after upgrading from 6.1.0 to 6.1.2 this past week.

Also seeing tons of these errors being generated via SYSLOG from the LDR's experiencing issues. And these errors are only showing up on two of the LDR's having the issue.

routing[822]: PROBLEM 0x3d02-42 (0000):

According to support they have zero knowledge of this error code. Still trying to dig deeper and figure this out.

[SpasKaloferov]

HI,

can you give me the name of the support person or the case ID so that i can follow up with him on this.

BR,
Spas Kaloferov

[admin]

mrlesmithjr wrote:

@tanagi - Were you able to sort this out? I myself have been going through a very similar issue since around 12.20.2014 and still seeing the same issues after upgrading from 6.1.0 to 6.1.2 this past week.

Also seeing tons of these errors being generated via SYSLOG from the LDR's experiencing issues. And these errors are only showing up on two of the LDR's having the issue.

routing[822]: PROBLEM 0x3d02-42 (0000):

According to support they have zero knowledge of this error code. Still trying to dig deeper and figure this out.

Could you please provide a bit more info around your situation, or a support ticket number?

[dlizotte]

This sounds very familiar to an issue that we had run across in our lab when we went through an upgrade from 6.0.4 to 6.1 and then to 6.1.1. This upgrade was all performed in a single day so there were no changes to config between 6.1 and 6.1.1. The problem manifested itself as being unable to ping through an LDR and we went through numerous attempts to repair this problem. The steps we took were to rebuild hosts, discovered a NIC driver issue with our servers so upgraded that, verified MTU sizes for vxlan transport which was ok, Removed and re-installed controllers. Tore down and rebuilt all logical infrastructure within the environment. None of these attempts made any difference, we also opened a case with VMware to attempt to remedy the solution but due to timelines we actaully ended rebuilding our lab environment back to 6.0.4 and performed an upgrade directly to 6.1.2 from 6.0.4 and this seems to have resolved this issue. I am not saying that there were problems for sure in the 6.1 version but it appears that the common point to all of these problems are that upgrading to 6.1 and then another upgrade from there seems to create this same problem.