Upgrade to 2.5G. Question about Adaptive Load Balancing, USB Hub, best setup

[GiamBoscaro]

My provider is upgrading my connection to 2.5G and giving me a new router with 2.5G ports, so it is finally time to upgrade my DS423+ connection to 2.5G.
I will buy an USB ethernet adapter and install the drivers. I had already tested this in the past and I was able to make everything work, but I did not have then any support for the 2.5G in my home network.
Anyway, I do have few questions regarding my setup:

1. I will keep the 1G LAN connection active as a backup, in case the 2.5G drivers fail. Does Synology automatically gives the 2.5G priority and then switches to the 1G in case of problems?
2. At the moment I am using both 1G LAN port with the Adaptive Load Balancing. Is it worth to keep this or should I just keep a single LAN connection?
3. At the moment I am using both the USB ports in the Synology. One is for the UPS data connection, one for an always connected 2.5" drive for the backups. Now I will be using one USB port for the ethernet adapter, thus I will need to use a USB Hub to connect both the HDD and the UPS to the other USB port of the Synology. Do you think I could have some problems with this configuration? Anything that I should be aware of?
4. Is there a difference in using the front or back USB port for the ethernet adapter?

[bb-qq]

Due to Synology system limitations, achieving reliable redundancy with BOND will likely prove challenging.
Instead, assigning different IP addresses to each NIC and configuring the internal NIC as a backup is simpler, requires less consideration, and allows for a more cost-effective environment setup.
Note that in this case, Synology's NICs must be assigned different subnet addresses. Assigning the same subnet makes it unpredictable which NIC will be used, negating the performance benefits.

Regarding USB, the front ports tend to be more stable, though this appears to depend on the environment.
Since Ethernet adapters are the least error-tolerant and consume the most power, connecting the UPS and HDD via a USB hub is a good idea.

[GiamBoscaro]

Due to Synology system limitations, achieving reliable redundancy with BOND will likely prove challenging. Instead, assigning different IP addresses to each NIC and configuring the internal NIC as a backup is simpler, requires less consideration, and allows for a more cost-effective environment setup. Note that in this case, Synology's NICs must be assigned different subnet addresses. Assigning the same subnet makes it unpredictable which NIC will be used, negating the performance benefits.

Regarding USB, the front ports tend to be more stable, though this appears to depend on the environment. Since Ethernet adapters are the least error-tolerant and consume the most power, connecting the UPS and HDD via a USB hub is a good idea.

Ok so in summary:

• Front USB for the adapter ✅
• Disable LAN bonding ✅
• 1G LAN set as backup ✅

What I did not understand is: Synology's NICs must be assigned different subnet addresses. Do you mean that the 2.5 Adapter should have a different subnet than the two 1G LAN, or do you mean that the two 1G ports should have a different subnet?

At this point, I would probably use just one of the 1G ports. Does it really make sense to have both of them connected and as a backup?

If the 2.5G adapter breaks, then I need to use the 1G LAN ports, but these are in a different subnet thus I will not be able to connect to them remotely. How do you handle failures then?
My idea was that Synology would automatically switch to the 1G LAN, having the router port forward another port to that LAN as a backup connection that I can only use, or having another device in the same network with Tailscale. But again this means also being in the same subnet as the router.

Do people use some script maybe to check if the 2.5G adapter is working, and maybe tries to restart the drivers or even the NAS? Could this work? Are there some examples of this?

[bb-qq]

I wrote the previous post assuming the use of one built-in NIC and one USB NIC.

DS423+
- builtin NIC: 192.168.2.1
- USB NIC: 192.168.1.1

PC
- NIC: 192.168.1.1 (and 192.168.2.1 as alias)

In this case, the system won't automatically switch to the built-in NIC if the USB NIC fails, but you can still perform actions like remotely rebooting the system. This should be sufficient in most cases. Using Tailscale is a bit more complex, but you can achieve the same remote management by setting up a jump host and assigning two addresses to it.

I am not aware of any examples of monitoring NICs using scripts. It seems achievable using cron scripts.

[GiamBoscaro]

I wrote the previous post assuming the use of one built-in NIC and one USB NIC.

DS423+
- builtin NIC: 192.168.2.1
- USB NIC: 192.168.1.1

PC
- NIC: 192.168.1.1 (and 192.168.2.1 as alias)

In this case, the system won't automatically switch to the built-in NIC if the USB NIC fails, but you can still perform actions like remotely rebooting the system. This should be sufficient in most cases. Using Tailscale is a bit more complex, but you can achieve the same remote management by setting up a jump host and assigning two addresses to it.

I am not aware of any examples of monitoring NICs using scripts. It seems achievable using cron scripts.

What I do not understand is, that if the USB NIC fails, then the NAS will be connected to the network 192.168.2.1. But my gateway is at 192.168.1.1. The network are different, so can the NAS actually communicate with the router at 192.168.1.1 and connect to the internet?
I do not have any networking experience more than I need, but from what I know the NAS will be isolated. If I was at home I could switch subnet on my workstation and ssh into the Synology to restart, but remotely?
In my mind using a device with Tailscale as a jump server, with a NIC aliased to the other subnet, could work. Remote connection through my router, I don't understand how it could work.