The core issue stems from the operating system's routing table. When Docker creates its internal bridge networks, these networks typically register routes on the host (in this case, the vSAN File Services appliances).
If the internal Docker network range overlaps with an external network segment ###.##.#.#/16 production subnet, the operating system's routing table might prioritize the internal Docker route over the route to the external physical network interface.
Consequently, traffic destined for the production ###.##.#.#/16 subnet is misrouted towards the internal Docker bridge instead of the physical network where the production subnet resides, leading to a "black hole" ( region where data packets are silently discarded without notifying the sender) for that traffic.
To avoid this routing conflict, the internal Docker network range used by vSAN File Services appliances must be explicitly changed to a non-overlapping private IP range before activating vSAN File Services.
The procedure involves using the `configure-vsan-file-service` command-line utility on the vCenter Server Appliance (VCSA) *before* you enable vSAN File Services in the vCenter UI.
/16. While /16 is a common default for Docker bridges, you can configure it to be a smaller subnet, such as a /24, /20, or any other appropriate private CIDR block, as long as it does not overlap with any of your existing production networks or other internal Docker networks./usr/lib/vmware-vsan/bin/configure-vsan-file-service --set-container-network 0.0.0.0/0 before activating vSAN File Services?
/usr/lib/vmware-vsan/bin/configure-vsan-file-service is indeed the correct utility. However, you should NOT use 0.0.0.0/0 as the CIDR. 0.0.0.0/0 represents all IP addresses and would cause massive routing issues. You must specify a specific, non-overlapping private CIDR block that you have chosen for the internal Docker network (e.g., ##.###.#.#/16, ###.###.###.#/24).