Router replacement to hub 2 from hub 1

[#1]

Hi,

Router firmware: 4.7.5.1.83.8.289.1.3

Please can I request a replacement router. I am experiencing DNS issues every day. The line is mostly stable - I run monitoring on a home server and there are no connection issues (low latency and no packet loss) apart from the odd blip, with the exception of DNS.

All WiFi goes through a UniFi network. All devices report above 95% WiFi experience scores, so no issues with WiFi.

UniFi access points are reporting the following DNS anomalies every day:

Date	Events	Timeouts	Latency	Devices Affected
Mar 21	89	48	41	8
Mar 22	85	27	58	7

Users constantly find slow or no connections, but WiFi stays up. Router WAN logs show no disconnects, and monitoring shows constant uptime (ping to 1.1.1.1). Occasionally users are redirected to the router homepage when loading websites.

The issue is intermittent. Here are results from dig at three separate times showing the router's DNS relay is failing. All three servers are queried simultaneously from the same device on the same connection:

Commands:

dig @192.168.1.254 google.com   # Hub One DNS relay
dig @212.159.6.9 google.com     # Plusnet DNS direct
dig @8.8.8.8 google.com         # Google DNS

Test 1:

                 Query 1   / Query 2   / Query 3
Hub One relay:   1,638ms   / 655ms     / 2,062ms
Plusnet direct:  12ms      / 13ms      / 12ms
Google direct:   12ms      / 13ms      / 12ms

Test 2:

                 Query 1   / Query 2   / Query 3
Hub One relay:   TIMEOUT   / 1,497ms   / 2,003ms
Plusnet direct:  14ms      / 14ms      / 19ms
Google direct:   14ms      / 14ms      / 19ms

Test 3:

                 Query 1   / Query 2   / Query 3
Hub One relay:   3,387ms   / TIMEOUT   / 2,241ms
Plusnet direct:  12ms      / 12ms      / 12ms
Google direct:   12ms      / 14ms      / 14ms

[#2]

@Agreed4403 Admittedly, you have an 'old' Hub, but I don't quite understand your testing parameters, especially the 'Hub1 relay' one. Can you elaborate?

[#3]

The tests show three “dig” queries to google.com sent at the same time, each routed through a different DNS resolver so the results are directly comparable under identical network conditions:
Hub One relay (192.168.1.254) - the router’s built-in DNS relay, which is what every device on the network uses by default via DHCP

Plusnet DNS direct (212.159.6.9) - Plusnet’s upstream DNS server, queried directly, bypassing the Hub entirely

Google DNS direct (8.8.8.8) - a neutral third-party reference point, also bypassing the Hub


Since all three queries travel over the same line at the same time, the connection itself is not a variable. Plusnet and Google consistently respond in 12-19ms across all nine queries. The Hub One relay by contrast ranges from 655ms to 3,387ms and times out completely on two occasions.
This isolates the problem to the Hub One’s DNS relay process. The upstream servers and the line are both healthy.​​​​​​​​​​​​​​​​

[#4]

OK, thanks, that explains what you are doing.

I admit to not being an expert in this area, but I didn't experience these issues when I was on PN (ADSL).

I may be barking up the entirely wrong tree here, sorry if I am, but what do trace routes to the affected sites show?

[#5]

@Agreed4403 

Have you considered buying your own modem/router?

Be warned the Hub Two only has three LAN ports, the WiFi names can only share the same name, which might not be an issue.

With DNS there is the option to use your preferred resolver.

Also the Hub two can be set to bridge mode (modem only device).

You might need to phone in and commit to a new contract to get a Hub two.

[#6]

😀 Glad you asked, since the issues started I have been logging a traceroute to 8.8.8.8 (Google) every 6 mins and that has been pretty much perfect, apart from the odd blip, usually with high wind which has always been the case on this line.

Here are the logs from the traceroute from the times I queried via 'dig':

Test 1 — 0% loss, 19.7ms latency
Test 2 — 0% loss, 13.5ms latency
Test 3 — 0% loss, 13.7ms latency

i would add that there isn't a specific affected site, this effects all domains.

Every time we have suffered the dropouts and lag, the traceroute has shown the connection has been fine, so the issue must be with domains being resolved.

[#7]

I would but as I understand I would also need to buy a separate VDSL modem. I already have 3 UniFi APs and switches. Also I'm holding out for FTTP and hope I would get a decent router supplied then.

Hub Two would be fine. At the moment with the Hub One, WiFi is turned off and only 2 LAN ports are used to 2 switches.

Yep, can find no option to set DNS on the Hub One. Of course I can set it at the device which has fixed the issue, but there are a lot of IoT devices that I can't set DNS on, plus 6 people live here, multiple devices, guests... it's a lot to keep setting DNS for everyone.

[#8]

---

@Agreed4403 wrote:

I would but as I understand I would also need to buy a separate VDSL modem.

---

I appreciate you are replying to @Dan_the_Van , but the last time I looked, admittedly a while back, there were plenty of combined modem/routers still on the market.

[#9]

@Agreed4403 

As @jab1  says there are VDSL modem routers available new, there is always the option of buying a used product.

The current router offering for Full Fibre is the Hub two so you might be depending on FF product consider an improved router which supports WiFi 6 and above.

I do recommend phoning plusnet.