Critical error in Hub while in Modem Mode

Looking at the Hub's Upstream and Downstream status tabs do you see signal levels that are out of the preferred range or many errors being logged ?

We need the Upstream as well, thanks.  Also setup a BQM to monitor the incoming circuit from VM  http://www.thinkbroadband.com/ping

What is supplying the network at the moment if the Hub is in Modem mode?  Is this device(s) working as expected?

You should not need to reboot the hub if you can copy the  TP Link Deco WAN MAC to your PC NIC then connect it the hub in place of the TP Link Deco to see if it connects fine.

This is still only part of the available info.
We have not seen the 3.0 Downstream power levels or any of the 3.1 Down or Up stats.

So 18 T3 timeout affecting all of the 3.0 upstream channels, if that does not change I'd suspect an engineer making a connection at a local street cabinet. It is quite unusual to see identical count of T3 error on all channels.

Same for the Downstream Post RS errors, if they now remain stable the issue has passed.

I have had an engineer to the house, and replaced the original hub5.
They could see no issue.

How would I go about escalating and refering to this thread?

Hub 3 replaced with a Hub 5 . . . 

This might be a clue,  if the TP Link Deco is connected to the 2.5GB/s port of the Hub 5, the TP Link Deco must have Energy Efficient Ethernet aka EEE Disabled on its WAN port for stable & reliable operation.

If this is not possible, the TP Link Deco should be connected to a 1GB/s port of the Hub 5.  

If you're seeing the error “No Ranging Response received – T3 time-out”, which is a DOCSIS error indicating your modem is struggling to communicate properly with VM’s network equipment (specifically, the CMTS). This usually means that even though the modem sends a Ranging Request, it doesn’t get a response, causing it to reset the connection. 

This error often points to issues in the physical connection typically on the upstream path. This could be caused by:

• Faulty, corroded, or loosely connected coax cables/connector
• Poor-quality or malfunctioning splitters/amplifiers
• Interference or noise upstream
• Issues outside your property, such as worn or overloaded infrastructure (taps, nodes, ports, etc)

If you report this became apparent after upgrading to a Hub5, The Hub3 uses DOCSIS 3.0 while the Hub5 runs on DOCSIS 3.1. On paper, DOCSIS 3.1 is a big upgrade. It allows more capacity and higher speeds but it’s also more sensitive to any noise or signal imperfections on the line. With DOCSIS 3.0 (Hub3), the connection tends to be more forgiving, so you could have borderline signal levels (e.g. upstream power slightly high, or downstream SNR a bit low) and still not notice many issues. With DOCSIS 3.1 (Hub5), those same borderline levels can cause daily drop-outs, packet loss, or latency spikes.

Or your TP Link Deco is sending out ARP and DHCP of your LAN from many source MAC and Docsis disconnects you till a reboot unlikely but who could say?

Test with a PC to the hub in modem mode and see if it disconnects.