Intermittent connection issues loading/changing site- secure connection failed/site can't be reached

[EDIflyer]

Is anyone else occasionally getting a 'connection down' error from Chrome (or 'Secure Connection Failed' from Firefox) when changing page - if I hit refresh it works fine. At first I wondered if it was the site I was using but have noticed it on multiple different major websites over the past couple of months and not noticed it from other locations where I don't use Three Broadband so does seem to be related to the router/connection (this is via wired Ethernet too, so not a WiFi problem).  I've tried changing DNS server in case that helps but no difference.

It mainly seems to happen when trying to first load a site or (annoyingly) at checkout when a different site is being loaded as part of the checkout process.  It certainly doesn't happen everytime but does happen with reasonable frequency.  I've also noticed I often get it when trying to pull/push from/to Github too and have to do so a number of times for it to work (browsing the Github website works fine).

I'm using a NR5103E with Firmware Version V1.00(ACBJ.0)b14 - I tried rebooting it but to no effect.

[AkiTaiyo]

After about a month of being absolutely fine.. the EOF errors have returned today after another short outage.

I spoke to a customer support rep a while back before the problem stopped, they initially told me that I'd need to pay off the remaining contract, but I said did that still apply when Three can't deliver a service I'm paying for.  After that they said they could cancel without charge if I returned the router. 

I've been trailing EE at the same time whilst this has been happening, turns out they are also on the same mast so still get good reception and speed.   Their service has worked perfectly but a little more expensive to make the switch permanent (I've been trialing a PAYG sim). 
I didn't make the switch in the end since the errors stopped happening for me, but now they are back I think I'll look at doing it again.

[fashric]

The complete non-response to this issue has convinced me that when my contract is up, I definitely won't be renewing.

[bytespider]

I'm going to cancel mine too. I've already suggested to several friends and family members not to go with Three. I'd like to hear some other options from others that are considering cancelling or have cancelled and moved to another provider.

[toaster]

I've been doing some more investigations. My tests have been running daily. I won't list them all out but in summary I get failures on a number of sites - typically:

- Microsoft.com - around 3-5%

- My test https server (AWS) - around 2%

-wordpress.org - around 25-30%

- youtube.com, duckduckgo.com, amazon.com, s3.amasonaws.com - around 0.5-1.0%

I was away from home for half term which allowed me to test a few things.

My phone contract is with a MVNO that uses the Three network. As I was travelling I was able to run tests in multiple areas of the country (i.e. different masts) and compare the results against my home network (Three) and local fibre internet (2 different networks).

I found the following:

- I can see the same pattern of behaviour with the MVNO and the Three network regardless of where I'm connected to. The suggests that it's not restricted to a single mast and that MVNOs that use the Three network have similar issues (This may be an interesting avenue to escalate this issue).

- I ran a test at home (Three network) and at the same time the same test using fibre networks. I saw 0 errors on the fibre networks. I think this rules out an issue with the websites themselves.

- When travelling, I used my phone as the router rather than the NR5103E. I think this rules out it being an issue with the router (this supports previous investigations in this thread)

[Midnight54]

Not been running any tests today but after switched off work vpn man its been incredibly bad today especially - sometimes taking half a dozen refreshes to load many pages.  @JonathanB  Come on man, why no updates on this

[Midnight54]

Ran it for a few hours tonight, (bearing in mind this is the 5000 rather than 500 (I'm assuming in millisecond) delay between each try

Once more today, Real-debrid is showing around 40% failure, Santander approx 5% (cant be bothered to work exact figures), 

And yet more days without any response from support on what action is being took

For those that volunteered to do testing for them, did any engineers actually get back to you?

[toaster]

I was asked to confirm what times I was available but the engineering team never confirmed. @JonathanB said they'd try to rearrange for the following week but I haven't heard anything - that was about 2 weeks ago.

[bytespider]

Nope, Nada, Not a Sausage. Thoroughly disappointed 

[Midnight54]

With AkiTaiyo's help, modified the other script slightly and ran it last several minutes, ssl errors being logged successfullly now.  Reddit for most part returns failures for some reason, but still gets SSL errors once in a while.  I suspect that if I increase the delay between each check that the SSL errors might increase

Mostly been working over VPN today as had lot of work to do so not tested much of actual real life browsing, but the issue is definitely still there

[Midnight54]

Increased the RequestInterval from 500 to 5000 to slow down between checks, and real-debrid is significantly worse for SSL failures right now - with 21 out of 44 requests SSL erroring, and 9 out of 44 so far for Santander