For some time now the VM DNS Servers have had an annoying issue where they periodically stop responding to new requests. It's got worse recently and now happens numerous times a day.
The symptoms are thus: Any existing web browser or application sessions (e.g. Skype) continue to work, but attempting to open a new connection requiring DNS resolution will fail. This affects web browsers, Skype, email, streaming, etc. In short, anything that requires a DNS lookup.
Sometimes this will self-correct after a few minutes, but often it lasts 30 minutes or more and on some occasions the only way to resolve it is to re-start the router.
Changing the DNS Server settings manually to another DNS provider (e.g. Google Public DNS on 22.214.171.124 / 126.96.36.199) fixes the problem immediately.
My guess is that many of the folk reporting dropped broadband connections on here are actually experiencing DNS failures not actual connection failure.
[For reference, I am using a SuperHub2 in modem mode, with a WNDR3700 router on VIVID200]
Thanks for the input. I have too many devices connected to do it on a per device basis (and not all of them are WinX with HOSTS file access). Changing the router DNS Server settings is easy, quick and fixes the problem which gives me back a decently reliable connection which I what I need.
I'm not going to hold my breath waiting for VM to fix a problem with what is after all a fundamental part of their system, as it took them best part of a year to fix the over utilisation problems locally, but there must be a ton of other folk out there who are affected by this but for whom any mention of configuring manual DNS settings will result in the response "manual what?" and that means they will likely be stuck with a sub-standard service...
Apologies for your connection issues. To be honest I'm doubtful that this is a DNS issue per se as there is a localised fault affecting your broadband connection. A network engineer has been assigned to investigate and we'll keep you updated as soon as he reports back.
But the fault (ref: F005188613) is for low SNR and FEC's which will cause speed loss and disconnections your side.
I'll take a look tomorrow to see if we have any progress news, though the network segment, when I tested it just now, was looking fairly error free.
@Jen_A Thanks for the info on the fault. I have noticed one or two brief actual outages recently, but generally my Power Levels, SNR, etc. look fine (within 'Seth's Reasoned Levels'), although the Hub2 Log has a lot of "No Ranging Response received - T3 time-out" errors. There are fluctuations and speed is down a bit from 'normal', but that's not the issue which is affecting the DNS - see below.
@ravenstar68 Thanks for the info, but I'm not using Chrome (I'm a Firefox person ).
The problem I am seeing is that everything can be OK for hours or even days and then DNS lookup fails. It affects all devices connected to the router, whether wireless or wired. This may be for a few minutes or last longer so that the only sensible solution is to re-start my router after which normal operation returns.
During these times, the connection itself does not drop. Any sites that are already open in a web browser or App (e.g. Facebook) continue to work and can be browsed freely (as the browser or App already has an open connection and knows the IP address of the site). It's only when trying to open a new site that the problem arises, as the browser or App makes a DNS request which gets no response.
Since changing the DNS servers to 188.8.131.52 / 184.108.40.206 the problem appears (apart from the odd connection glitch, probably due to the fault) to have disappeared. It's too early to say that that's definitive as not enough time has elapsed since I made the change, but I'm certainly not seeing the daily multiple occurrences of the issue that I was with the VM servers.
Coincidentally, I have a VM Tech coming to fit a Hub3 tomorrow (23 Mar) as I've upped my broadband service to VIVID200 Gamer (to get the faster upload speed, NOT to play games!!). I will update further after the new setup has been installed and given a while to settle...
Sorry for not getting back to you last Thursday. But the SNR fault was closed on that day, the 22nd, in the afternoon.
If you do have any further issues with your connection just let us know and we'll take a look at the upstream segment. With this type of fault there is always a possibility that it could recur, more so if the errors abated before the actual fault source could be found.
Thanks for the update. The connection has been error free for several days now.
However, since I first flagged the issue, the SNR fault has been (hopefully) fixed, my Hub has been changed to a Hub3 and I changed the DNS server settings to use the Google Public DNS, rather than the VM defaults.
As you are no doubt well aware, ideally fault isolation is best if only one thing at a time changes - changing 3 things as has now happened makes it more difficult to determine the original cause. I will give it a few more days as-is to establish a new baseline and then revert the DNS server settings in my router to the normal VM ones and see how it goes.
I agree with you, three 'fixes' has muddied the waters somewhat with regards to ascertaining what the cause was. I still think it was the SNR fault but your idea sounds good. If you change your DNS servers whilst the connection is stable we can then see if there are any further problems for you. As I've said, if you post an update to notify us of any change I (or one of my colleagues) can check the network for you.