vSAN Health Service - Network - Storage cluster client network connectivity check

Products

• VMware vSAN

Issue/Introduction

This article introduces the new vSAN Health Service check 'Storage cluster client network connectivity check' in vSAN 9.1 and provides details on why it might report an error and how to fix the error state.

This health check is called 'vSAN Max Client Network connectivity check' in vSAN 9.0.

Environment

VMware vSAN 9.0 

VMware vSAN 9.1

Resolution

Q: What does the 'Storage Cluster Client Network connectivity check' check do? 

It performs a connectivity check for Storage Cluster Client Network by checking the heartbeats from each host to all other hosts in the server cluster. If this check fails, there is a basic issue with network connectivity for Storage cluster client Network.

Q: What does it mean when it is in an error state?

The remote client cluster might run with degraded performance. It may impact the objects accessibility from the remote client cluster when all interfaces for Storage cluster client network can't connect to each other. 

The health check will be shown as yellow (warning) when the Storage cluster client network connectivity is unstable among ESXi hosts and if the Storage cluster client network connectivity is down among ESXi hosts, which could indicate either virtual network configuration issues (such as problems with VMkernel adapters and Virtual Switches), or physical network issues (such as problems with cables, physical NICs and physical switches), a red health is displayed.

Q: How does one troubleshoot and fix the error state?

Check for packet loss over the Storage cluster client network by running the vmkping command from source ESXi host to the destination host's IP address of the Storage cluster client network. 'Storage cluster client VMkernel NIC' is the VMkernel adapter which is enabled Storage cluster client. 'Destination IP' is the IP address of VMkernel adapter which is enabled the Storage cluster client from the target host.

Command: vmkping -I <vSAN Storage cluster Client vmk> <vSAN host IP>

Refer to Troubleshooting the vSAN Network for more details.

If there's no obvious networking issue, dump the missing heartbeat count from the source host with the below command and contact Broadcom for support.

Command: /bin/python -c "import pyCMMDS;output=pyCMMDS.GetExtVmkNicDetails(<vSAN Cluster UUID>);print(output)"