Google and Friends Release Net Neutrality Measuring Tools

angry tapir writes "Google and a group of partners have released a set of tools designed to help broadband customers and researchers measure performance of Internet connections. The set of tools, at MeasurementLab.net, includes a network diagnostic tool, a network path diagnostic tool and a tool to measure whether the user's broadband provider is slowing BitTorrent peer-to-peer (P-to-P) traffic. Coming soon to the M-Lab applications is a tool to determine whether a broadband provider is giving some traffic a lower priority than other traffic, and a tool to determine whether a provider is degrading certain users or applications. 'Transparency is our goal,' said Vint Cerf, chief Internet evangelist at Google and a co-developer of TCP/IP. 'Our intent is to make more [information] visible for all who are interested in the way the network is functioning at all layers.'"

Comments

God bless em

[nametaken]

This is great, I'm sending this link around to friends and family on different networks now.

Not because I want to know, but because I want them to read it, see their problems in b&w, and be aware of what their ISP's are doing... without me preaching to the deaf.

Re:God bless em

[Zartan the Great]

Who posts a Bandwidth Tester on Slashdot and then expects us all to be able to use it at once... Cheers

Re:God bless em

[davester666]

After RTFS, my first thought is that all the major ISP's will reverse engineer the tools, such that their traffic 'bandwidth shaping' methods will actually prioritize these packets, so that end users wind up getting lied to (that their network traffic isn't being slowed down AND that they are getting a faster internet connection than they actually are).

Re:God bless em

[lysergic.acid]

those tools seem pretty useful, but i don't know how user-friendly some of them are. personally, i'm looking for a tool to see if our ISP (at the office) is hijacking our DNS errors, or all of our computers are just infected with malware.

also, is anyone else seeing a bunch of "" characters on the Network Diagnostic Tester homepage? is my browser/system screwed up, or are there a bunch of a little boxes with "FF FD" in them scattered all over the page?

Comcast

[Chabo]

As a Comcast customer, I heartily look forward to trying out these tools on my Comcastic(tm) connection at home!

On another note, I also look forward to carrier-grade NAT in the near future, when Comcast decides they want to stay with IPv4 forever!

Define slowing

[thule]

What do they mean by slowing? You can "slow" Bittorrent by shaping or by giving it less priority? Again, is this being confused on purpose? To what end? From my post on the Cox story:

One issue is over subscription. Unless a company is large enough to have lots and lots of peer connections, your ISP is probably over subscribes their upstream connections. This is fine, because on average traffic goes in bursts. The problem is that everything starts to break down once you have a small pool of people running P2P 24/7. These people are just as greedy as the ISP's they complain about. They want a huge "dedicated" pipe, but have others subsidize it. I have no issue with someone like Cox de-prioritizing their traffic so that the people that just want their Vonage to work don't get squashed out. This is a temporary solution because the ISP will eventually have to up their pipe speed.

The other issue is granting certain companies privileges on a network and penalizing other companies they don't like (e.g. penalize Vonage and prioritize a VoIP partner). This should be illegal. This is a clear case of violation of neutrality. At the same time, the company should be able to directly peer with a company (say a VoIP provider) without violating the law. This may seem unfair, but peering has been a perfectly valid way of reducing traffic on a transit connection.

The last issue is traffic caps. I don't think there should be a law against it as long as the company is upfront about it. Putting caps on traffic allows ISP's to maximize their over subscription and cater to people that want low cost Internet service. We *want* people to afford Internet services. The market chooses. If you are a big user of P2P, then you will have to go with another ISP that does not have caps. You may have to pay more for this privilege... sorry, but that is how things go. The market must have a way to manage scarcity of resources. If you want more of a resource, you will have to pay for it even it if looks the same (e.g. 5mbit from Cox versus 5mbit from FiOS).

Don't confuse QoS with net neutrality. As long as the QoS is applied equally, then it should be perfectly fine.

Re:Define slowing

[tkrotchko]

"Putting caps on traffic allows ISP's to maximize their over subscription and cater to people that want low cost Internet service."

I realize you're talking in theory.

But now that Comcast has capped traffic, have they provided a new, inexpensive tier of service? Or has their prices gone up? Can you name any company that capped traffic and then lowered prices?

Seems to me capping traffic is simply a way to stay revenue neutral, and reduce costs. Let's not pretend it's for any other reason.

Re:Define slowing

[Anonymous Coward]

"Go with someone else"

Because that is *totally* possible in Comcast markets.

Re:Define slowing

[riceboy50]

I think the more sane of those arguing against throttling are those who point to the ISPs' advertising, which is often at best misleading and at worst fraudulent.

Re:Define slowing

[Ichijo]

I have no issue with someone like Cox de-prioritizing their traffic so that the people that just want their Vonage to work don't get squashed out.

Why deprioritize at all? Give everyone using the pipe at a given moment an equal portion of the available bandwidth. Divide it up evenly by customer, not by application. One person doing p2p shouldn't affect another person's Vonage phone call or vice-versa.

Re:Define slowing

[isBandGeek()]

The ISPs have large areas where they are the only high-speed Internet providers, besides expensive and high-ping satellite connections. You know just as well as I that there's no feasible way to build your own ISP. Caps are only possible because of the ISPs' anti-competitive behavior. What you're saying is, "Hey, DeBeers has a business model of 'managing costs'. They can do what they want. If you don't like it, find another player (never mind that DeBeers controls 90%+ of the market) or make your own diamond mining corporation."

Re:Define slowing

[lysergic.acid]

so it's greedy to expect an ISP to deliver to you the service they advertised and that you've paid for? don't confuse your own solipsism & selfishness with other people's being greed. right now you're saying that VoIP should have priority over P2P because presumably "ordinary" people like you use VoIP but don't use P2P (a rather questionable assumption). so just because someone else's internet usage patterns are different from yours, your traffic should be given priority over theirs, even though you both pay the same monthly rate?

you also seem to be the one confusing the issue of file-sharing with so-called "bandwidth hogs." first of all, congratulations on buying into (or trying to perpetuate) the ISP's scapegoating of power users and file-sharers for their poor service--i'm sure all those Asian countries with cheap, symmetric high-speed broadband connections don't have file sharers or power users. secondly, even if we assume that a broadband provider has to oversell in order to remain profitable (an unlikely case), why could a simple bandwidth cap be implemented regardless of the type of traffic one has? protocol discrimination and deep packet analysis (which simply adds more network overhead) is not necessary even if you're trying to perform damage control after having over-sold by too much.

at our office i use BitTorrent maybe once a month to download 30-40 MB Photoshop brush sets, or an 18 MB Ad-Aware install file (the LavaSoft site requires you to sign up for Trialplay, and give out your personal information and CC# to get the Anniversary edition), and only very occasionally an up-to-date Windows XP disc image (700~800MB). on average, our monthly BitTorrent traffic totals less than 100MB on a 10Mbps connection.

on the other hand, we're a record label so we listen to band demos all day long, and these days most of it is done via MySpace, which is very convenient; we can see how many plays each artist has received that day, what shows they've played recently, and just gauge their popularity more easily. it also cuts down on the demo CDs being pressed/burnt/shipped, which is good for the environment. however, this means we're streaming music all day long (from 9 AM to 5 PM). assuming the average audio quality frm myspace is 96kbps, that's about 330MB of traffic from streaming audio alone, not to mention all the banners, photos, and other graphics on these bands' MySpace pages.

so if 2 people each consume, say, 500~600MB of network bandwidth each day, but one person uses it solely for BitTorrent while the other uses it solely for sending large files via e-mail, why should the BitTorrent user's network packets have lower priority than the e-mail user? how is he being greedy or asking others to subsidize his bandwidth?

ISPs have no business dictating how a broadband subscriber uses his internet connection. if they want to throttle people's connections after a bandwidth cap is exceeded, fine--don't advertise the service as unlimited, make the cap clear to your customers, and apply it equally to everyone regardless of whether they're an old grandma who's watching the Food Network in HD on her cable TV, or if it's a teenager downloading the latest Slackware ISO via BitTorrent.

lastly, if an ISP cannot meet the demands of their customers, they need to do one of two things: a.) upgrade their infrastructure to increase network capacity, or b.) don't oversell so much. the basic concept of overselling is sound. on average not everyone is going to use 100% of their pipe 100% of the time. but it's up to the ISP to calculate what their average network usage is going to be, and provide enough total network bandwidth so that the network doesn't become saturated during peak hours. what you don't do is try to scapegoat power users for your own miscalculations and continue to oversell while trying to dictate how the public uses the internet.

most countries are offering faster broadband at lower costs, following the usage trends that are shifting towards high bandwidth applications

Sincere-O-meter

[carrier lost]

The set of tools, at MeasurementLab.net, includes a network diagnostic tool, a network path diagnostic tool and a tool to measure whether the user's broadband provider is slowing BitTorrent peer-to-peer (P-to-P) traffic.

Will there be a tool to tell me if Digital Max is really my friend in the digital world, or if he's just bullshitting me?

Useful to both p2p users and network admins.

[fenix849]

These tools are no doubt going to be very useful to everyone that uses p2p software for _any_ purpose.

The flipside is that as an administrator of a workplace network i can also use these tools to ascertain whether or not the traffic managment and qos i've put in place on the corporate network is working.

It doesn't really matter so much on this particular network as p2p protocols are blocked (infact every outgoing port is blocked from the internal lan, some https sites are whitelisted, and all non-ssl web access is proxied.

But it will allow me to ensure the qos for our voip trunks is effective.

This Glastnost thing ain't gonna work

[morgan_greywolf]

It got Slashdotted:

caused by: java.io.IOException: open HTTP connection failed.
        at sun.applet.AppletClassLoader.getBytes(AppletClassLoader.java:265)
        at sun.applet.AppletClassLoader.access$100(AppletClassLoader.java:43)
        at sun.applet.AppletClassLoader$1.run(AppletClassLoader.java:152)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.applet.AppletClassLoader.findClass(AppletClassLoader.java:149) ... 9 more

ComObjectCast and Friends Fire Back

[Raystonn]

"ComObjectCast and a group of partners have released a set of tools designed to help broadband providers and researchers determine the algorithms used by Net Neutrality Measuring Tools. The set of tools, at MeasurementLabSucks.net, includes an enduser diagnostic tool, an enduser pathfinding diagnostic tool, and a tool to determine is the enduser is measuring whether the user's broadband provider is slowing BitTorrent peer-to-peer (P-to-P) traffic. Coming soon to the M-Lab-Sucks applications is a tool to determine whether an enduser is using a tool to determine that a broadband provider is giving some traffic a lower priority than other traffic, and a tool to determine if an enduser is using a tool to determine whether a provider is degrading certain users or applications. 'Obfuscation is our goal,' said Argle-bargle GlypfpGlopf, Chief obfuscation evangelist at ComObjectCast and a co-developer of ROFL/MAO. 'Our intent is to make more [information] visible for all who are interested in keeping customers from using what they paid for.'"