Web connection timing out

[Rbbhghs]

1. Hi there,

My connection seems to have become problematic over the last month or two. (Timeframe difficult to pinpoint).

ive made the following adjustments to try to eliminate issues but it persists:

 - plugged into master socket test socket

 - swapped EE route rain for router which worked well with BT Broadband previously 

 - wired and wireless connection 

we live about 200 yards from the exchange and are syncing at full speed:

Type:

  Upstream Downstream

Current Rate (kbps)	19999	75215
Max Rate (kbps)	22084	77869
SNR Margin (dB)	6.5	6.3
Line Attenuation (dB)	18.3	20
Errors (pkts)	8	0

Pings come back at 100pc from 8.8.8.8 and speed is great, but the web is unusable - I just get timeout of the vast majority of connections initiated, irrespective of site (this site included, which I’m accessing over 4g so I can make this post).

does anyone have any idea what’s going on? It’s like the web is being filtered and the proxy is overloaded, or similar...

[XRaySpeX]

What is timing out? Is it the BB disconnecting or is it your devices disconnecting from the router while the BB is still up? Are you using a wireless or wired connection?

If it's the BB & your EE router is a Brightbox that will tell us more info.

[Rbbhghs]

Hi there,

across both the EE box and the other router I have, across /all devices/ web connections are timing out intermittently.

action is: open web page, wait, then times out and page not available showing.

The broadband remains up and whilst I  have issues on one device, it seems like I don’t have them on another at the same time - ie - it seems like it’s specific to the device connection, not the BB connection.

im also unable to do video calls and voip calls without them dropping, but I’ve not tested this thoroughly as I’ve just been moving to 4g to avoid issues.

[XRaySpeX]

You could try & get round this issue by switching the DNS in the router to Manual with 1 of these public DNS:

OpenDNS   :	208.67.222.222 & 208.67.220.220
Google DNS:	8.8.8.8 & 8.8.4.4 
Cloudflare DNS	1.1.1.1 & 1.0.0.1 
Norton DNS:	198.153.192.40 & 198.153.194.40

[Rbbhghs]

Hi there,

Sorry, i'm already using 8.8.8.8 and 8.8.8.4 as my DNS.

I pretty sure it's not DNS as my video and phone calls are also dropping.

By way of example, right now i'm on a conference call on Zoom. This is connected over a combination of TCP and UDP connections direct to IP addresses, with no DNS used over my EE 4G connection (which ironically Speedtest says is faster than my broadband..(!) https://www.speedtest.net/my-result/i/3028755938 vs http://www.speedtest.net/result/7927847477 )

.

When i connect my phone to the wifi, the connection is dropped for about 10 seconds every so often. Immediately as soon as this happens, i'm able to then pick up my laptop  and access google and search without issue, but then 2 minutes later the web connection will drop off my laptop whilst the phone connection remains stable. I'd expect it to be a connection issue with the internet connection, not the devices themselves, but i'm sat right next to the router and i've no issues at all with the wifi. 

It's very odd.

What's annoying is that having being troublesome on my call, it's now been fine for the last 10 mins so i'm not able to do more tests...

i spent 5 years working debugging enterprise IT network issues so i'm used to the usual battery of things you'd do here, but i'm completely at a loss. The closest i can theorise is an upstream proxy or firewall running out of sockets and therefore the client is shutting down the connection ungracefully, but even that wouldn't account for my voip connection dropping as i'd expect that to be maintaining an open connection, not trying to open a new one...

[Rbbhghs]

By way of sharing another symptom, on Amazon Prime Video, this is fairly typical.

I'm not sure how the page is built inside the app, but i'm guessing it's a series of REST web services over an SSL connection. There is probably a loop that iterates around each section in the page making a call each time. Some work, some don't...

The Prime app handles this slightly more elegantly by failing the individual service call, not the whole page, so the user experience is degraded, but the app doesn't crash.