单ESXi主机环境下通过vSphere Lifecycle Manager Images更新组件的相关技术疑问

Let's walk through your questions one by one—this is a super common scenario for single-host remote or homelab setups, so we'll break it down clearly:

1. vCenter自身VM关闭后，更新如何执行？ #

When you start remediation via vSphere Lifecycle Manager (LCM), vCenter orchestrates the entire process upfront before any VMs go down:

• First, it pushes all necessary update payloads and execution commands to the ESXi host. This means the host already has everything it needs to complete the update without vCenter being online.
• Next, it will shut down your specified VMs (including vCenter and the router VM).
• The host then enters maintenance mode, applies the driver update, and reboots.
• Once the host comes back online, it will automatically power on your VMs in the order you've configured in the host's VM startup settings.

Critical tip: Make sure you set your router VM to start before the vCenter VM in the host's startup configuration. This ensures vCenter has network connectivity as soon as it boots up—otherwise, it might come online but fail to connect until the router starts, causing delays or post-update issues.

2. 更新是否会自动启动？是否需要用户交互？ #

Once you kick off the remediation process (after configuring the right options), the entire workflow is fully automated. You won't need any manual input during the update—perfect for remote servers where KVM access is costly or unavailable.

Just confirm these options are enabled in your remediation settings:

• Auto-shutdown of VMs
• Auto-enter maintenance mode
• Auto-reboot after update completion
• Auto-power on VMs post-reboot

With these set, the host will handle everything from shutting down VMs to bringing them back up once the update is done.

3. 使用LCM Images还是esxcli software vib install更合适？ #

Both methods work, but there are key tradeoffs to consider:

• vSphere Lifecycle Manager Images: This is VMware's recommended approach moving forward (since VUM is being deprecated in vCenter 9). It manages your host's entire software stack as a single, consistent image, making it easier to track components, ensure compliance, and apply future updates seamlessly. If you import all your manufacturer-provided VIBs/offline bundles into LCM and add them to your host image, you can keep your configuration standardized.
• esxcli software vib install: This manual approach is often recommended by hardware vendors because it gives direct control over individual VIB installations. It might be necessary if the vendor hasn't tested their VIBs with LCM, or if there are specific dependency rules LCM doesn't handle.

Since you plan to install multiple VIBs from the same manufacturer, here's a suggested workflow:

1. Import all vendor-provided VIBs/offline bundles into LCM.
2. Add them all to your host's custom image.
3. Run the pre-check again to catch any compatibility conflicts.
4. Proceed with remediation if the pre-check passes.

If the pre-check flags issues, fall back to the vendor's esxcli method: manually put the host in maintenance mode, shut down all VMs, install each VIB via esxcli software vib install -v /path/to/vib, then reboot the host. Either way, ensure the 4600 smartpqi driver is installed first if the other VIBs depend on it.

4. Quick Boot不支持但已启用，是否需要禁用？ #

Absolutely disable Quick Boot. The pre-check explicitly states it's not supported on your host, so enabling it can cause unexpected problems—like failed updates, hardware initialization errors, or even the host failing to boot back up.

In the remediation options, uncheck the "Enable Quick Boot" box before starting the process. This ensures the host performs a full reboot, which is necessary for the driver update to take effect properly on unsupported hardware.

备注：内容来源于stack exchange，提问作者Danny Beckett