Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Unsure what exactly has caused it but it looks like one of VM's connections to LINX is running over capacity.

Traffic has been rebalanced between the links, see if this improves things more tonight :)

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Have they not upgraded to 100Gb yet?

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Depends if you include the 100Gb UPC / Liberty Global have.

Virgin / ntl have a long history of carrying considerably less LINX capacity than their peers, preferring to peer in more locations and have more entry and exit points on their network.

Many haven't upgraded to 100Gb to be fair. There are only about 25 100GbE ports lit on LINX. Akamai alone have 6 of them, Apple have 2, Facebook have 2, Microsoft 2.

ISP-wise BT have 2, Sky have 3, TalkTalk have none: they use 32 x 10Gb.

VM use 23 x 10Gb on the London LANs. Their network pushes >4Tb/s at peak so they must have a ton of other public exchanges, private peering and transit taking the strain.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Well that depends if they're using it to carry VM broadband traffic :p:

That's true, but VM also have several times higher average speeds than Sky and BT, and BT for that matter have had 100Gb LINX ports since at least 2011. EE also have two. Basically everyone bigger than VM has multiple, despite VM having higher average usage per customer.

Unfortunately I see that as part of the problem, as it's much quicker to add public peering capacity than new private peerings when dealing with unexpected traffic increases.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

It is but then causes issues as far as how you actually get all that traffic robustly and resiliently to that public peering point.

Comcast, etc, run the same model for pretty good reasons. VM could ram all 4Tb/s through London but not a good idea. It may be quicker but it's unwise, far better to try and get traffic off your network as soon as is feasible.

It's a different equation for Sky as the heritage of their network is the old Easynet network and a massively fibre dense ring in London and for xDSL ISPs as they can centralise IP traffic and use BT Wholesale as a transport network but for networks like Virgin's where customer traffic is clear IP very close to the customer makes sense to have a bunch of exit points.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

The evening bump is back for me!

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Yep, spectacularly too - after being almost eliminated for a couple of days

http://www.thinkbroadband.com/ping/s...05-06-2015.png

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Cardiff Friday I notice the close similarity to the previous poster's graph

http://www.thinkbroadband.com/ping/s...06-06-2015.png

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

mine is also spiking around 10pm and has done for around 2 weeks.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

The peak-time hump should, hopefully, be sorted now.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

My min latency now has a step in it.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Cant do anything online with my connection atm gaming wise, and skype is awful for me :S

http://www.thinkbroadband.com/ping/s...12-06-2015.png


Thats with nothing in the background running either no utorrent or anything.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Could one of you who's more clued up about this than me tell me if this is good or bad?

https://www.cableforum.co.uk/images/local/2015/06/8.png

Been running the monitor for a few days now on our connection and it generally seems to follow a similar pattern to the above in the previous days as well. This is from the SH2ac directly at the moment, not in modem mode. Whether that's what's causing the issue and a better router would fix it I'm not sure.

I would have thought the latency spikes would be when the connection is getting used, but they seem to be fairly regularly spaced, even through the very early morning when the only devices connected to the router are phones/tablets in standby.

I haven't noticed any connection drops or slowness and generally I'm happy with the connection so far, but that graph does seem to have more regular latency spikes than some that have been posted here.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

http://www.thinkbroadband.com/ping/s...07-06-2015.png

The above is a SH2 in router mode.

The below is a SH2ac in modem mode on the same VM segment.

https://www.cableforum.co.uk/images/...2016/02/30.png

So there is a defo difference in BQM behaviour as between router and modem mode but not to the extent of the spikes shown in the dead of night on your circuit.

Re: Think Broadband Ping Monitor Results (POST YOURS)
 

Some routers are a little more 'casual' about responding to pings on the WAN than others :)

EDIT: What are the first 3 octets of the IP addresses each of those have?

Mind tracerouting from both to 80.249.99.164 please?