Intermittent connectivity issues have been observed between source cloud VMs and the HCX NE-R appliance on the target site when NSX/DLR is configured at the source site in conjunction with a cloud-based NSX DLR.

Products

Issue/Introduction

• This is only applicable for customers using NSX/DLR at source in addition to cloud NSX DLR
• This is applicable from HCX 4.10.0 onwards where the API changes were introduced.
• When the source cloud VM and the NE-R appliance are hosted on the same ESXi host, communication may work fine or behave inconsistently and may not always function as expected.

Environment

This issue is only applicable to customers using NSX/DLR at the source site in conjunction with a cloud-based NSX DLR. Customers running Distributed Virtual Switch (DVS) extensions with a non-NSX setup at the source will not be affected.

Cause

This is a known issue applicable from HCX version 4.10.0 onwards.

Resolution

This issue has been fixed and verified in HCX versions 4.10.2, 4.11.0, and all later versions.

Additional Information

Temporary workaround: Unextend/Re-extend the same segment.

Note: whenever app-engine service will be restarted or HCX Cloud Manager rebooted, then this issue will be seen again.

Sustainable Workaround: Enabling MON on the extended segment will prevent this issue from occurring, as it allows configuration of specific properties at the port level for each VM on the logical segment. The source or remote gateway can still be retained as the preferred gateway for all L3 routing purposes.