mudmanc4

That moment when you can breath, between the KVM cutting out fsck letting you know your fkd, then finally this:

No, your correct this is not looking good........ at all for this disk

Here, join the fun:

Running fsck is always a sooper blast, so while I'm waiting .........................

How do you say fsck, as in explaining it to someone what you are doing.

Myself, fsuck

I'm all for secure transmission of data, and the EFF appears to be on the right side of a lot, even offering free certs.

But this 'thing' they are doing with browser extensions. Not so sure about all that.

Chuck the addon to the bin and see what happens.

Update?

26 minutes ago, CA3LE said:

And the 3.7% difference between what you set in pfSense (1000 Kbps) and your TMN speed test result (964 Kbps) can be accounted for in overhead.

 

TCP Over IP Bandwidth Overhead

I was going to write that in the original but couldn't prove it, but yea!

Let this sink in for a moment.

Now have a look at setting in squid locally:

FreeBSD

1 hour ago, nodsa said:

@mudmanc4

 

am now running alternate tests with an older laptop runing ubuntu live cd..........id in results is garage, it appears that winblows is restricting upload speeds by around  5mbps ??

That would not shock me. But why is this happening.

Software firewall ? Antivirus? Anything secondary you might have running on the machine?

In that case, I would surmise you as in the VDSL modem) is roughly  1km away from the cabinet, which would account for the ~17% loss

So all in all, your doing good. If in fact the distance is correct, reverse calculation and assumption on average results compared to standard deviation.

What about wiring inside your residence, age / quality, splitters, filters ect. All of which will account for signal strength.

As for the modem ( ~2-3 years old) the appliance has likely has several updates since placement, this happens transparently via ISP.  We'd have to know the firmware, version compared to new subscribers device / firmware revision and even reasoning, to know if that device plays any part in deviation.

If you click "Results" in the header of the page, the link will take you to https://testmy.net/quickstats/nodsa There you can delete your own records.

The stats graph in your first post, I placed there for your and mine reference, it is all tests taken from your ID, logged in or not. Which I believe is drawn from the unique net device MAC

The test appliance, or script talktalk uses will live on their own network, so testing will define your basic connectivity and throughput to the ISP

Can I ask what 'speeds' or package you are paying for?

If you are consistently or on average achieving 65/13 , I would say that is decent, not optimal. This would coincide with what the ISP is selling you

@jjgk , Nice connection!

What do your real world results look like from/to various locations?

Compare these 4Gig CentOS file downloads :

Located in California

http://mirrors.kernel.org/centos/7/isos/x86_64/CentOS-7-x86_64-DVD-1708.iso

Located in Delaware:

http://mirror.lug.udel.edu/pub/centos/7/isos/x86_64/CentOS-7-x86_64-DVD-1708.iso

Located in Florida:

http://mirror.atlantic.net/centos/7/isos/x86_64/CentOS-7-x86_64-DVD-1708.iso

Located in Ohio:

http://mirror.cisp.com/CentOS/7/isos/x86_64/CentOS-7-x86_64-DVD-1708.iso

Located in Texas:

http://yum.tamu.edu/centos/7/isos/x86_64/CentOS-7-x86_64-DVD-1708.iso

@nodsa , Is the image a screenshot from talktalk or elsewhere.

2 hours ago, nodsa said:

I see the minimum guaranteed speed as an absolute figure, which will be maintained by my ISP,

Exactly, to add, To the ISP , after in which is completely out of ISP hands. In other words ISP's as a general rule internationally, take their (in my opinion) collaborative responsibility to maintain  functional and fluent means to access to grant data via end user (which consists of all of us) , and toss it to the wind.

Why call yourself an Internet Service Provider , if in fact your unable to support such claims in a reasonable manner, not that of which you define or clause in a customer agreement to pay.

Lets look at it this way. The system has been built backwards. From the very start, we should have built major infrastructure interconnecting major cities, then spider outward.

Hind site is great I know, but were still doing the same thing. Adding nodes to the 'internet' without the support of connectivity.

I've since setup squid, not that this will help with streaming, but every little bit will, caching 500Gigs should do something over long term. Each user gets 750k .

I've yet to work out whether squid will do throttling if :80 / :443 are not explicitly blocked on the network, or if traffic shaper will continue to do it.

Here's a quick screen of streaming youtube @ 480p, (where 'thecave' is a local domain) also notice the same net as yours? I know what they are, but I don't know what other than SSL through google this will be accomplishing so...........

drill -x 172.30.199.50
;; ->>HEADER<<- opcode: QUERY, rcode: NXDOMAIN, id: 44882
;; flags: qr aa rd ra ; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0 
;; QUESTION SECTION:
;; 50.199.30.172.in-addr.arpa.	IN	PTR

;; ANSWER SECTION:

;; AUTHORITY SECTION:
30.172.in-addr.arpa.	10800	IN	SOA	localhost. nobody.invalid. 1 3600 1200 604800 10800

;; ADDITIONAL SECTION:

;; Query time: 0 msec
;; SERVER: 10.10.5.1
;; WHEN: Fri Jan 19 11:34:31 2018
;; MSG SIZE  rcvd: 103

Interesting.

drill 172.23.0.49
;; ->>HEADER<<- opcode: QUERY, rcode: NXDOMAIN, id: 6202
;; flags: qr rd ra ; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0 
;; QUESTION SECTION:
;; 172.23.0.49.	IN	A

;; ANSWER SECTION:

;; AUTHORITY SECTION:
.	3395	IN	SOA	a.root-servers.net. nstld.verisign-grs.com. 2018011900 1800 900 604800 86400

;; ADDITIONAL SECTION:

;; Query time: 73 msec
;; SERVER: 10.10.5.1
;; WHEN: Fri Jan 19 11:29:05 2018
;; MSG SIZE  rcvd: 104

As well as :

drill 172.30.199.50
;; ->>HEADER<<- opcode: QUERY, rcode: NXDOMAIN, id: 30320
;; flags: qr rd ra ; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0 
;; QUESTION SECTION:
;; 172.30.199.50.	IN	A

;; ANSWER SECTION:

;; AUTHORITY SECTION:
.	3569	IN	SOA	a.root-servers.net. nstld.verisign-grs.com. 2018011900 1800 900 604800 86400

;; ADDITIONAL SECTION:

;; Query time: 29 msec
;; SERVER: 10.10.5.1
;; WHEN: Fri Jan 19 11:32:44 2018
;; MSG SIZE  rcvd: 106

Over the years reading posts, we could find more statements such as yours than not. Tech comes out, everything is awesome, tech leaves, soon thereafter, connection stinks on ice.

Aside that, tech was likely testing on the ISP level network, never touching the 'internet' , or anything outside of their control.

Also, check out the mirrors testmy.net uses, find one closer to you and run some testing

Ouch, yea that penalty is rough.

Have you run a tcpdump, Is the media streaming from within the ISP network, much as netflix does?

The limits I stated are set by myself on one vlan, I've since loosened them up a touch, was rather useless as it was.

I've since added an additional 1000ms to each 100 packets coming through the pipe

And dropped 1 out of every 100 packets.

Anyone thinking how much of a dick move this is, in the last ~6 weeks plus, I've not watched one movie, giving all I can, and still ignored.

Time for a lesson in reality.

Ya know what I should be doing , is setting it up 56k old school, like we had to go through. That would however result in server timeouts I'm certain.

Hell, the limiter is set at 500kbps now, where it's really not feasible to load much, but will stream youtube somehow.

~10-12+ gigs /day got us here

ISP charges $100 per overage , that would be the last couple days of the month. Silly rabbit, just bit off your toes

And this is what it looks like on my vlan with a vpn running

This is what it looks like on my vlan