After rebooting a vCenter Server Appliance (VCSA), the Direct Console User Interface (DCUI) displays the error message: "no compatible network adapter found". This error prevents users from configuring or monitoring the network settings directly from the console. When checking the VCSA's network configuration via SSH or console, the primary network interface (vmxnet3) is found to be named eno1 (or similar, like ensXXX) instead of the traditional eth0. Despite eno1 potentially having a valid IP address and allowing SSH/VAMI (Appliance Management Interface) access, the DCUI specifically reports the absence of a compatible network adapter. This behavior typically arises after maintenance activities, such as taking offline snapshots, applying updates, or a general reboot, where previously the vCenter Server functioned without issue.
VMware vCenter Server Appliance (VCSA)
Dell VxRail
The underlying operating system for the vCenter Server Appliance, Photon OS (like many modern Linux distributions), utilizes Predictable Network Interface Names (PNIN). During the boot process, a udev rule (e.g., /usr/lib/udev/rules.d/60-net.rules) and associated systemd.link configurations are designed to automatically rename network interfaces based on their hardware characteristics (such as PCI bus ID or MAC address). Specifically, the vmxnet3 virtual network adapter, which is initially identified by the Linux kernel as eth0, is subsequently renamed to a predictable name like eno1 by these rules. The vCenter Server Appliance's DCUI, however, is often developed with a hardcoded dependency or expectation for the traditional eth0 interface name. When the DCUI fails to find an interface explicitly named eth0, it reports the "no compatible network adapter found" error, regardless of whether network connectivity is actually available through the renamed eno1 interface. This is a common conflict between modern Linux interface naming conventions and application-specific expectations.
To resolve this, the vmxnet3 network adapter must be consistently named eth0 by overriding the default predictable naming rules. The recommended and most robust approach for Photon OS is to use a systemd.link file, which is processed early in the boot sequence.
eth0 NamingAlt + F1 (or Alt + F2) to switch to a shell prompt. Log in as root.eno1.ip a show eno1
link/ether line (e.g., link/ether XX:XX:XX:XX:XX:XX). Note down this MAC address (XX:XX:XX:XX:YY:ZZ) as it is required for the next step.systemd.link File:
10- prefix ensures this rule is processed very early, taking precedence over default naming schemes.vi /etc/systemd/network/10-persistent-eth0.link
XX:XX:XX:XX:XX:XX with the actual MAC address obtained in Step 2.
[Match]
MACAddress=XX:XX:XX:XX:XX:XX # IMPORTANT: Replace with the actual MAC address
[Link]
Name=eth0
vi editor (:wq).10-eth0.network:
/etc/systemd/network/10-eth0.network exists and is correctly configured for your vCenter's IP address (static or DHCP). This file should automatically apply to eth0 once the naming is fixed. If this file was modified or deleted, it might need to be recreated or corrected via vami_config_net or the VAMI web interface.cat /etc/systemd/network/10-eth0.network
[Match]
Name=eth0
[Network]
Address=###.###.###.###/24 # Your vCenter IP and CIDR
Gateway=###.###.###.### # Your Gateway IP
DNS=###.###.###.### # Your DNS Server
Domains=########.local # Optional
[Match]
Name=eth0
[Network]
DHCP=ipv4
systemctl daemon-reload
reboot
eth0 with its network configuration.ip a. Confirm that the primary network interface is now named eth0 and has the correct IP address.nslookup <your_vcenter_hostname> to verify DNS resolution.service-control --status --all
https://<vCenter_IP>:5480 and log in as root. Check the "Summary" tab for overall health status (CPU, Memory, Storage, Database, Swap).https://<vCenter_IP>) to confirm vCenter connectivity, host status, and the ability to perform basic operations.