Community Forum

Data Usage Meter Launched

Xfinity Forum Archive
About the archive project

Xfinity Forum Archive...

This is an archived section of the community.

Content in this area has been identified as outdated or irrelevant.

This change was done in an effort to make the forum easier to use and to keep only the most helpful and recent content active.

Post your questions in the Xfinity Community

Contributor

Re: Data Usage Meter Launched


jlivingood wrote:
Thanks for advising where you are.  Now that I'm back at work after the hoilidays, I need to start working through the deployment schedule.  I'll advise when I have more info.

 

Any chance you've gotten time to look at the deployment schedule?
New Poster

Re: Big Brother

Indeed - this is just a setup for rate increases. And you can kiss that 250 gig limit goodbye as that is just an arbitrary number at this point. Never thought I would have to consider DSL or wireless but this just about does it for me. I signed on for unlimited internet period. I download zip but I do not like the concept at all. What next, Comcast starts sending our surfing info to Barry O. and the nannies in DC?
Frequent Visitor

Re: Data Usage Meter Launched

Yes, there is one device listed.  It looks like a MAC address.
Official Employee

Re: Data Usage Meter Launched


ObjetDart wrote:

Hi,

 

I was excited to receive the "Usage Meter: Pilot Launch" email over 2 weeks ago, but when I follow the instructions and go to "Users & Settings" like it says, it just says "The usage meter is not available for this account." 

 

I don't understand why I would receive this email if I'm not part of the pilot.

 

I am located in Eugene, OR.

 

Thanks!


Do you see any devices there?  If not, contact customer care (e-chat recommended) and ask them to fix your account (incorrect device provisoning).

JL
Internet Services
Silver Problem Solver

Re: Data Usage Meter Launched

If Jason (jlivingood in the post above yours) doesn't reply, send him a personal message by clicking on his name. 

 

Frequent Visitor

Re: Data Usage Meter Launched

Hi,

 

I was excited to receive the "Usage Meter: Pilot Launch" email over 2 weeks ago, but when I follow the instructions and go to "Users & Settings" like it says, it just says "The usage meter is not available for this account." 

 

I don't understand why I would receive this email if I'm not part of the pilot.

 

I am located in Eugene, OR.

 

Thanks!

Official Employee

Re: Data Usage Meter Launched

Thanks for advising where you are.  Now that I'm back at work after the hoilidays, I need to start working through the deployment schedule.  I'll advise when I have more info.
JL
Internet Services
Frequent Visitor

Re: Data Usage Meter Launched


jlivingood wrote:

dh7132 wrote:

I'll believe the meter when I see it.


What market are you located in?


Chambersburg PA

Contributor

Re: Data Usage Meter Launched

We'll happily try all markets.  I'll inquire for Chicago (Northern 'Burbs).
Official Employee

Re: Data Usage Meter Launched


dh7132 wrote:

I'll believe the meter when I see it.


What market are you located in?

JL
Internet Services
Frequent Visitor

Re: Data Usage Meter Launched

Well, I'll say this again. (I'm sure I'm wrong)Smiley Wink

I believe the delay of the usage meter allows enough time for Comcast to "ferret out" and attempt to get rid of the heaviest users, thus allowing more bandwidth for several more customers.

Bandwidth = limited.  Customers = unlimited.  More $

Centurylink DSL has no cap.

 

I'd be curious to know how many customers have been axed since the limit's been imposed.

I'll believe the meter when I see it.

Bronze Problem Solver

Re: Data Usage Meter Launched


flabbergast wrote:

Bethkatz,

 

Well met.    I can see your point, however; my interpretation of net neutrality differs in some ways.  

 

You see, I tend to think of net neutrality as having three main goals:  content independence, the end of data rate limiting and the elimination of what TM and Comcast have ironically deemed "Fair Sharing" of bandwidth.  

 

Well, the fact that you don't understand the concept of Net Neutrality is neither here nor there.

 

I find it ironic because there really isn't anything fair about selling your customers a service without telling them up front what they are being sold (telling them in size 2 font on the last page of an SLA is a sleazy way to get a customer to "accept" certain terms).

 

You have a point there. On the other hand, it's extremely unlikely that anyone who would be affected by a 250 MB cap would be unaware of its existence.

 

One question that comes to mind is why Qwest can offer a completely unregulated broadband experience while Comcast finds it necessary to constrain its customers' service with rate limits, bandwidth limits, and content filtering?    

 

Another question that comes to mind is: Why do you make up stuff?

 

http://www.qwest.com/internethelp/eup/16915_EUP_Details-15.pdf 

 

What content is Comcast filtering?

 

If Comcast's service was superior, one may argue that you'd be paying a premium for premium services.  However, like everyone knows, cable internet is about as stable as a 21 year old actress living in hollywood.  I see my service fluctuate from great to horrid in an hourly fashion.   For example, earlier today I was getting about 3Mbps down with pings to google around 150ms.   An hour later, it was back up to a sustained 7Mbps with pings to google around 50ms.

 

You do know that Comcast doesn't host Google, right? 

 

What that tells me is that Comcast does not want to invest in their network.  

 

That doesn't even tell you where the problem was. BTW, my standard Comcast service level has gone from 1.5 Mbs to 8+ Mbs over the last few years. 

 

My hope is that net neutrality will force ISPs to compete once again and when given a choice, most customers will select an unregulated service over a regulated one.  

 

Almost all customers would choose a service with a 250 GB cap rather than a more expensive uncapped service. 

 

As it stands, Comcast is taking advantage of its customers and as an telecom professional, it makes me sad to see what the industry has become.  

 

Get out the crying towel, then. You can have caps on flat rate service or you can have service tiers, and you're gonna be unhappy about both options, aren't you?

 

 

Regular Contributor

Re: Data Usage Meter Launched

The internet is intrinsically rate limited as it is a shared resource utilizing limited capacity equipment. As to "fair sharing" that doesn't mean free. Comcast charges more for faster connections (Mb/second). What they haven't done yet is charge more for more data (Mb/month). Net neutrality is about not giving xyx company better access to comcast customers because XYZ company pays Comcast for better access. It isn't that an ISP can't limit speed or volume, only that the ISP can't limit traffic from some sources and not others. Or limit a particular type of traffic while not limiting other types of traffic at the same time. I wonder if that means they can't refuse email from known SPAM sources? That is one place where Comcast places a significant (read total) limit on access. Net neutrality does permit Comcast to balance traffic to its customers that are sharing a limited resource such that one customer can't unduly monopolize the shared resource. What Comcast got called on was restricting usage for P2P traffic strictly because it was P2P traffic.

 

Before you question my comment about not being able to stop SPAM remember that government regulations are a case study in the "law of unintended consequences". 

Silver Problem Solver

Re: Data Usage Meter Launched

Comcast has a separate mechanism in place to deal with that neighbor who's slowing things down for everyone else. See http://customer.comcast.com/Pages/FAQViewer.aspx?seoid=Frequently-Asked-Questions-about-Network-Mana... for an overview of it. Basically, if a node is overloaded, the heaviest users are given lower priority, so that the lighter users are not impacted.

 

This has nothing to do with net neutrality. Net neutrality is about not allowing ISPs to give preferential treatment to different content providers, e.g. Google can't pay Comcast so they'll give Youtube traffic higher priority than Hulu.

Problem Solver

Re: Data Usage Meter Launched

I'm not an internet expert and I don't really understand all of the fine points of net neutrality, but let me tell you what I think "fair" is in the context of being a Comcast customer.

 

If I have a neighbor up the street that downloads 250GB/mo of stuff around the clock and it slows down my service, then that is not fair to me.  I don't care if it is due to a network design decision that Comcast made years ago or for any other reason.  If there is someone hogging a limited resource that impacts the rest of us, then I expect something to be done about it. 

 

One thing that could be done would be for Comcast to spend unlimited $$ to build out the system so that everyone could download 1TB/mo, but that is not cost effective (and there would be someone in my neighborhood that would want to download 2TB/mo).  The other thing that could be done would be to ask the highest users to pay more so that Comcast could add the infrastructure to allow them to download huge amounts of data without impacting the rest of us.  That seems the most fair to me.

Silver Problem Solver

Re: Big Brother


BigBrotherBashe wrote:
I do not appreciate the "big brother meter" that was just announced.
Do you realize that the bandwidth cap is over a year old, the only thing that was just announced is a web page that will let you see where you stand relative to it? How can anyone not appreciate that?

 

Silver Problem Solver

Re: Data Usage Meter Launched

The 250 GB cap is not an absolute, permanent cap. Remember, to be considered an excessive user, you have to be above 250 GB and one of the heaviest users. When most customers are downloading movies every day, you'll have to be much heavier than 250 GB to become one of the heaviest users. If and when 250 GB becomes common, Comcast will have to raise this cap, since it would no longer be useful for them or us.

 

In any case, this thread is supposed to be about the bandwidth meter, not the bandwidth cap. Whether or not you consider 250 GB to be an appropriate cap, the meter is still a good thing, isn't it? 

Official Employee

Re: Data Usage Meter Launched

Replies inline

flabbergast wrote:
I see my service fluctuate from great to horrid in an hourly fashion.   For example, earlier today I was getting about 3Mbps down with pings to google around 150ms.   An hour later, it was back up to a sustained 7Mbps with pings to google around 50ms.  
[JL] Conditions on the Internet vary by time of day.  When it is peak hour, more people use the Internet and Google and other sites just naturally aren't as zippy when 1,000 people are accessing a server compared to 10 people. 

 

What that tells me is that Comcast does not want to invest in their network.  

 

[JL] Nothing in fact could be further from the truth.  Apart from initially pioneering 1st generation broadband, we continue to invest significant sums in our network.  You may have noticed our DOCSIS 3.0 deployment and the fact that we have been early and aggressive investors in D3.

 


 

JL
Internet Services
Official Employee

Re: Data Usage Meter Launched

 My replies inline in blue
 

 flabbergast wrote:

As a new member to the forum, I must wholeheartedly disagree with this statement.  

 

How can you tell what people are going to do with their connection?  Do you go to everyone's home and ask them politely, or perhaps take a ethernet capture, of what they are doing with the bandwidth that they pay for?  Probably not.  

 

[JL] Which is of course why a centralized data usage meter is worthwhile.

 

As we continue to enter the digital age, to say that 250GB of data is "a lot" is folly. I download game demos (1GB each), trailers, and HD movies (7GB each) quite often.  As Hulu and other streaming services become popular, the notion that ISP's can continue to limit a person's ability to gain access to these services will become obsolete per the statements made by Julius Genachowski, the FCC Chairman. 

 

[JL] Having a data usage limit, which limits excessive usage of a flat rate residential service, is by no means limiting anyone's ability to gain access to any services.  It is limiting excessive data usage.  Also, from the perspective of various regulatory groups, there is not a concern with having such a limit provided that it is clearly disclosed to customers and not a hidden or variable limit that is unevenly applied.

 

So I question the intent of Comcast to continually undermine not only the will of the customers, but the will of the FCC.  What is the intent here?

 

[JL] Nothing related to the meter is contrary to the direction the FCC is moving in or problematic from the standpoint of "network neutrality."

 

 In October of this year,  the FCC approved a notice of proposed rule making on the subject of net neutrality.  Essentially, sometime in 2010, ISPs will be forced, under penalty of law, to maintain unfettered access to the services that a customer as bought based on net neutrality.  This means no more 250GB limit, no more rate throttling and no more discriminating behavior.  

 

[JL] We don't engage in rate throttling or discriminatory behavior today, so nothing to change there.  Our transparency is now industry-leading (see http://networkmanagement.comcast.net), and we've been lauded for our protocol-agnostic congestion management system.

 

So again, I ask the question why Comcast continues to impose dictator-like rules and regulations on broadband customers?   It only serves to make Comcast appear as though they would rather sell smoke and mirrors to a customer than upgrade their networks to be more competitive in the market place   

 

I, for one, cannot wait for Comcast and other's to be put in their place by these new regulations.   Suppressing innovation and competition will only be tolerated for so long in this day and age. 

 

Chastise me if you must for my statements above.  Facts and evidence tend to have that affect on the irrational and incompetent individuals among us.   


 

JL
Internet Services
New Poster

Re: Data Usage Meter Launched

I love the way they make this sound like a good thing...what a bunch of **bleep**. We went from unlimited usage to a cap and somehow you market this as a feature. I hope this isn't the beginning of the end of Comcast.

 

If something affects 1% of users it is not a feature to punish the other 99% by adding a cap. No matter how you try and market it!

New Poster

Re: Data Usage Meter Launched

Bethkatz,

 

Well met.    I can see your point, however; my interpretation of net neutrality differs in some ways.  

 

You see, I tend to think of net neutrality as having three main goals:  content independence, the end of data rate limiting and the elimination of what TM and Comcast have ironically deemed "Fair Sharing" of bandwidth.  I find it ironic because there really isn't anything fair about selling your customers a service without telling them up front what they are being sold (telling them in size 2 font on the last page of an SLA is a sleazy way to get a customer to "accept" certain terms).

 

One question that comes to mind is why Qwest can offer a completely unregulated broadband experience while Comcast finds it necessary to constrain its customers' service with rate limits, bandwidth limits, and content filtering?    If Comcast's service was superior, one may argue that you'd be paying a premium for premium services.  However, like everyone knows, cable internet is about as stable as a 21 year old actress living in hollywood.  I see my service fluctuate from great to horrid in an hourly fashion.   For example, earlier today I was getting about 3Mbps down with pings to google around 150ms.   An hour later, it was back up to a sustained 7Mbps with pings to google around 50ms.  

 

What that tells me is that Comcast does not want to invest in their network.  My hope is that net neutrality will force ISPs to compete once again and when given a choice, most customers will select an unregulated service over a regulated one.  As it stands, Comcast is taking advantage of its customers and as an telecom professional, it makes me sad to see what the industry has become.  

 

Cheers.  

Silver Problem Solver

Re: Data Usage Meter Launched

Flabbergast, well said. There should be no chastising here for reasoned discussion. 

 

I agree with net neutrality in general except when quality of service is affected. We should have access to the Internet without preferential treatment for some data. The Wikipedia article on net neutrality is useful reading.

 

The data usage meter is counting bits and not considering content.

 

The 250GB limit will need to rise. But maybe that will become the basis of tiers. People who use more will pay more just as now people who want faster speeds pay more.

 

But having the meter lets us see how much we are really using rather than how much we think we are using. That will give us more facts and evidence. Comcast has stated what that there is currently a 250GB monthly limit and what they do if you exceed that. Now they are providing a meter to show you how much (not what) you are using.

 

I consider that a move forward to help us fight for fair access to the Internet. 

New Poster

Re: Data Usage Meter Launched

As a new member to the forum, I must wholeheartedly disagree with this statement.  

 

How can you tell what people are going to do with their connection?  Do you go to everyone's home and ask them politely, or perhaps take a ethernet capture, of what they are doing with the bandwidth that they pay for?  Probably not.  

 

As we continue to enter the digital age, to say that 250GB of data is "a lot" is folly. I download game demos (1GB each), trailers, and HD movies (7GB each) quite often.  As Hulu and other streaming services become popular, the notion that ISP's can continue to limit a person's ability to gain access to these services will become obsolete per the statements made by Julius Genachowski, the FCC Chairman. 

 

So I question the intent of Comcast to continually undermine not only the will of the customers, but the will of the FCC.  What is the intent here?

 

 In October of this year,  the FCC approved a notice of proposed rule making on the subject of net neutrality.  Essentially, sometime in 2010, ISPs will be forced, under penalty of law, to maintain unfettered access to the services that a customer as bought based on net neutrality.  This means no more 250GB limit, no more rate throttling and no more discriminating behavior.  

 

So again, I ask the question why Comcast continues to impose dictator-like rules and regulations on broadband customers?   It only serves to make Comcast appear as though they would rather sell smoke and mirrors to a customer than upgrade their networks to be more competitive in the market place   

 

I, for one, cannot wait for Comcast and other's to be put in their place by these new regulations.   Suppressing innovation and competition will only be tolerated for so long in this day and age. 

 

Chastise me if you must for my statements above.  Facts and evidence tend to have that affect on the irrational and incompetent individuals among us.   

Gold Problem Solver

Re: Big Brother

You should be happy to have a means of reassurance that you aren't among the 1%. 
New Poster

Re: Data Usage Meter Launched

I'm afraid I don't have a lot of content for you yet, but the general direction is excellent.  I know that I as a customer appreciate being able to see where I'm at on the bandwidth limit.  Good work, good direction. 
New Poster

Big Brother

I do not appreciate the "big brother meter" that was just announced.  Why attach this to everyone if "overusage does not apply to 99% of our customers"?  This service is VERY overpriced as it is..... stand by for a new fee schedule for us 99%ers..... It's time to shop elsewhere!
Bronze Problem Solver

Re: Data Usage Meter Launched

You do realize that the cap has been in place for over a year now, and that it's just the meter that is now being released? In other words, this has probably had no effect whatsoever on your service (nor on the service of about 98% of Comcast's subscriber base).
Message Edited by Joel on 12-17-2009 05:17 PM
New Poster

Re: Data Usage Meter Launched

Apparetnly it's time to start looking for another ISP. While it would take me quite some time to reach the 250GB cap, I did not sign up with comcast to be told how much bandwidth I can use. My cost keeps going up and the service keeps getting worse. Time to more on.
Silver Problem Solver

Re: Data Usage Meter Launched


laurie-beaverto wrote:
We are going with FIOS!
I could understand it if you were leaving because of the bandwidth cap, but it seems like you're complaining about the bandwidth meter instead. That seems backwards. Isn't the meter a good thing?

 

Problem Solver

Re: Data Usage Meter Launched


laurie-beaverto wrote:
We are going with FIOS!

Enjoy their usage meter!

 

Diamond Problem Solver

Re: Data Usage Meter Launched

CYA...
New Poster

Re: Data Usage Meter Launched

We are going with FIOS!
Silver Problem Solver

Re: Data Usage Meter Launched

Jason, thank you for the answers. I also have the secondary account although my husband rarely reads his Comcast email or does any account management.
Official Employee

Re: Data Usage Meter Launched

Inline.
 

bethkatz wrote:

I have a few more questions prompted by questions in the forum ...

 

1) Does the customer need to log in as the primary account holder to see the data usage meter? Or will it show up for the secondary account as well? It's my understanding that both accounts should show the same usage meter and that both should have access.

 

[JL] The primary acct holder can see the meter.  The secondary cannot, unless the primary gives them bill / acct level permissions.  That's pretty easy to do, and I do this on my acct, so my wife who uses a secondary has such access, FWIW.

 

2) Does the customer need to be connecting from their home and through the modem that's measured or can they check their usage from some other location? I'd think that it would be tied to the account. It should be tied to the account rather than physical location. Otherwise, you could be at someone's house, use their computer, and look at their data meter. That doesn't make sense. And I should be able to check the data usage of my home network from work once my area has the data meter deployed.

 

[JL] Anywhere that you can access Customer Central you can access the meter in cCentral.

 

3) When all the customers in the Portland area have the data meter available, could you please post that it has been completely deployed in that area?

 

[JL] Sure.  That should be next week sometime.

 


 

JL
Internet Services
Silver Problem Solver

Re: Data Usage Meter Launched

I have a few more questions prompted by questions in the forum ...

 

1) Does the customer need to log in as the primary account holder to see the data usage meter? Or will it show up for the secondary account as well? It's my understanding that both accounts should show the same usage meter and that both should have access.

 

2) Does the customer need to be connecting from their home and through the modem that's measured or can they check their usage from some other location? I'd think that it would be tied to the account. It should be tied to the account rather than physical location. Otherwise, you could be at someone's house, use their computer, and look at their data meter. That doesn't make sense. And I should be able to check the data usage of my home network from work once my area has the data meter deployed.

 

3) When all the customers in the Portland area have the data meter available, could you please post that it has been completely deployed in that area?

 

Contributor

Re: Data Usage Meter Launched

Got the email today. Finally!

 

Secondary account is still not synced with primary though so no device information shows up as well as voicemail on account. Whaaa...

Silver Problem Solver

Re: Data Usage Meter Launched

 
The limit was set in a time when people weren't downloading high-definition movies as a matter of course. If we see lots of users bumping into the limits, maybe we'll need to lobby for a higher limit. Or maybe it will be a non-issue. I have a feeling that Comcast doesn't really know what a realistic unobtrusive limit is.
 
The limit was established at a time when HD movie downloads were in fact being observed.  The limit is high and fair.  Compared to any other ISP with a limit that I am aware of in North America, 250GB is the highest, and in many cases by a long-run.
 


While this may have been done before, to try to put a little perspective on this.....
There are posts on BBR that indicate a good AVERAGE data transfer for a Netflix HD movie is APPROXIMATELY 1.5Gb (ignoring the 1000 vs 1024 bit arguments...notice I emphasize approximations here). Apparently, Netflix use VBR encoding, so different movies have different downoad sizes...
So, at 1.5Gb/movie, a quick back-of-the-envelope calculation says that to get close to your bandwidth limit, you would have to watch 6hrs of HD Netflix movies EVERY DAY for 30 days.
That's some pretty intense movie watching...... :

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I don't work for Comcast...


Help us to help you!!
- respond to requests for info
- post back if your issue is resolved
- mark appropriate posts as solutions


Community Icon
I am not a Comcast employee, I am a paying customer just like you!
I am an XFINITY Forum Expert and I am here to help. For information on the program click here.
We ask that you post publicly so people with similar questions may benefit from the conversation.
Was your question answered? Mark it as an accepted solution!solution Icon
Community Icon
I am not a Comcast employee. I am a paying customer just like you!
I am an XFINITY Forum Expert and I am here to help.
We ask that you post publicly so people with similar questions may benefit.
Was your question answered? Mark it as an accepted solution!solution Icon
Community Icon
I am not a Comcast employee.

Was your question answered?
Mark it as a solution!solution Icon

Contributor

Re: Data Usage Meter Launched


jlivingood wrote:

dukeofhurl wrote:
I am in Portland. I do not have access to the meter. Under My Devices it displays No device information is available.

More areas are being phased in on Monday - all of your area to complete by end of year.


 

Sounds good. Hope to get it today!
Bronze Problem Solver

Re: Data Usage Meter Launched


jlivingood wrote:
Replies in blue
The meter shows all traffic, both high-overhead and low-overhead, across the modem. So say for example I am a huge Internet TV and online movie user. Also assume these activities have an overhead rate of 20% of the traffic associated with these activities. Now lets assume you are a heavy FTP user and the overhead associated with that activity is 10%.
Keep in mind that overhead also includes packet retransmissions, also impossible to predict ahead of time and I have no idea how I'd account for that after the fact, with millions of users. 

You and I both access or send a total of 1GB of actual data content (payload) through our resprctive connections. My meter will show, ignoring rounding for this discussion, 1.2GB and your meter will show 1.1GB yet we both only accessed or sent exactly the same amount of content.

 

This is likely one reason the meter is so very high, as compared to other ISPs in the North American market, at 250GB instead of, say 20GB.  Oftentimes, these concerns are overblown when people see what their usage actually is.    

 

The report does indicate a good job with respect to the testing performed and it reports on the accuracy of the meter.

 

Thanks.  NetForecast had unprecedented access to everything we do.  I am not aware of any other ISP who has implemented a meter that has had a third party do such a study and release a public report.  We tested this meter for months upon months, and strove to be incredibly transparent about the subject.

 

I still think the process and procedures were inadequate to address the larger scheme of traffic monitoring, measurement, and reporting. It does not mater how accurate a meter is if it is not capturing and reporting the correct data.

 

The way we calculate data transfer is the way the entire (Internet) industry does so.  I am not aware of any company that does this differently, by somehow identifying and subtracting protocol traffic and the like.  I know you feel differently and all I can tell you is that this is how IPDR and other data transfers are measured - by just looking at the network interface and counting bytes moving across the interface.  That approach is simple, and it does not examine the contents of traffic.  To do it differently would likely require very invasive packet inspection that I would not be comfortable with, and it likely wouldn't scale particularly well.


 


Thanks again for responding. I personally am not concerned about the meter's implementation and as the report indicates it appears accurate. I just wanted to ensure I understood the testing process as well as the report and the rationale for its findings.

Official Employee

Re: Data Usage Meter Launched

Replies in blue
The meter shows all traffic, both high-overhead and low-overhead, across the modem. So say for example I am a huge Internet TV and online movie user. Also assume these activities have an overhead rate of 20% of the traffic associated with these activities. Now lets assume you are a heavy FTP user and the overhead associated with that activity is 10%.
Keep in mind that overhead also includes packet retransmissions, also impossible to predict ahead of time and I have no idea how I'd account for that after the fact, with millions of users. 

You and I both access or send a total of 1GB of actual data content (payload) through our resprctive connections. My meter will show, ignoring rounding for this discussion, 1.2GB and your meter will show 1.1GB yet we both only accessed or sent exactly the same amount of content.

 

This is likely one reason the meter is so very high, as compared to other ISPs in the North American market, at 250GB instead of, say 20GB.  Oftentimes, these concerns are overblown when people see what their usage actually is.    

 

The report does indicate a good job with respect to the testing performed and it reports on the accuracy of the meter.

 

Thanks.  NetForecast had unprecedented access to everything we do.  I am not aware of any other ISP who has implemented a meter that has had a third party do such a study and release a public report.  We tested this meter for months upon months, and strove to be incredibly transparent about the subject.

 

I still think the process and procedures were inadequate to address the larger scheme of traffic monitoring, measurement, and reporting. It does not mater how accurate a meter is if it is not capturing and reporting the correct data.

 

The way we calculate data transfer is the way the entire (Internet) industry does so.  I am not aware of any company that does this differently, by somehow identifying and subtracting protocol traffic and the like.  I know you feel differently and all I can tell you is that this is how IPDR and other data transfers are measured - by just looking at the network interface and counting bytes moving across the interface.  That approach is simple, and it does not examine the contents of traffic.  To do it differently would likely require very invasive packet inspection that I would not be comfortable with, and it likely wouldn't scale particularly well.


 

JL
Internet Services
Silver Problem Solver

Re: Data Usage Meter Launched

Jason, thank you very much for your answers. 

 

I expect that what we will see when this is rolled out to all of us is that the vast majority of us use a small fraction of the bandwidth we have available. 

Bronze Problem Solver

Re: Data Usage Meter Launched


jlivingood wrote:
Replies in blue.
 

The report explained a lot but I think it also raises questions.

 

1. The report describes 6.2% measured overhead as a result of the testing. This overhead is added to the meter. Testing consisted of transferring files using FTP which is a low overhead protocol. Why didn't the testers use a higher overhead protocol such as HTTP, which would be more representative of a typical Comcast users experience?

HTTP was not a suitable protocol for us to generate a huge volume of data transfer in a very short period of time.  Also, the point of the exercise was to try to minimize and control for the protocol overhead to the greatest extent possible so that we could more accurately gauge what the PC saw and then what each part of the measurement system recorded.  Thus, the goal was to measure measurement accuracy, and not to simulate the usage patterns of an "average user."  (Then we'd rathole on what average usage patterns are, etc., among other things.)  The other thing that is important to bear in mind is that many other protocols, and in particular HTTP, are highly variable in the overhead observed.  And so I didn't want to turn months of testing into years of testing.  If people have concerned over protocol overhead, I recommend participating at the IETF to make future protocols more data efficient (which often comes at the expense of functionality and extensibility).  ;-) 

 

2. The report goes on to identify several ways and devices that use the Internet that may be unknown to a user. The number of devices and their respective Internet usage will grow in time. There is no discussion in the report of how the bandwidth restrictions and monitoring will grow to accommodate growing use.

 

The NetForecast report is a study of the measurement system and it's accuracy.  It was not intended to address or examine Comcast policy changes, such as changing data usage policies.  However, the current system of using IPDR to collect usage statistics will continue into the future and will scale well, and it is built into DOCSIS 3.0.  I do not anticipate massive changes to the system, particularly as I am in the process of deploying it nationally after an extended development and testing period.  :-)

 

3. It is my understanding from the report that traffic tracking across the modem is captured and uniquely identified by the modem's ip address. The case where an ip address changes due to system maintenance, changed device connected to the modem, or other situation causing DHCP to assign a new ip addess is not addressed. What happens to the meter data when a new ip address is assigned? What happens to a user's account if the newly assigned ip address was one that during the month was previously assigned to a user who exceeded the 250GB limit? We've seen issues where ip addresses that were previously blocked were reassigned to subscribers.

 

IPDR is used to collect data usage statistics.  IPDR uses the modem's MAC address, and not the IP address.  As you indicate, the IP may change frequently, whereas the device MAC address does not change unless the device itself is changed. 

 

The report was good to read but it was incomplete in my opinion. I don't think the testing process and procedures adequately represented a typical Comcast user's traffic.

 

As indicated above, that was not a goal of the report.

 

It seems to me that traffic overhead should be a cost to Comcast of doing the business of providing Internet service. Measuring overhead and including this number in the usage meter is akin to FedEx charging $28 for a package then adding $1 for gas plus $1.50 for rent, plus ... Comcast IMHO should charge for the internet connection and not the overhead separately.

 

It is technically impossible in actual practice, and at reasonable expense, to separate out the protocol overhead from the actual data transfer.  The only way I could see to do it, which I would be strongly against, would be to deploy a massive (and I mean MASSIVE) Deep Packet Inspection system to examine the contents of all user traffic and perform content analysis and somehow separate out protocol-related traffic from data transfer.  This would involve a deep study of every protocol in existence and would be a moving target since protocol overhead, even for the same protocol, is HIGHLY variable and not fixed.  I simply don't see how such a system would be technically practicable or desirable.  Furthermore, it is not the state of the art to extract out protocol overhead when calculating data transfer.  It is understood that data transfer is inclusive of protocol communications, and this is how all data transfer measurement that I am aware of functions. 

 

Wait until they decide that Netflix users need to pay 10% overhead, HTTP users should pay 12% overhead, Internet TV viewers should pay 25% overhead, etc. Oh wait, that is exactly what this metering system does. If you are a high overhead protocol user you will reach the limit faster than a low overhead protocol user.

 

This is inaccurate and you are reading far too much into our statements about protocol overhead.  Without the exchange of protocol communications, not other communications can occur.  Protocol overhead is extraordinarily variable.  A very short communication, such an an IM, may have a lot of overhead relatively to the small payload of a message saying "Are you there?" (in such case, the protocol overhead would likely exceed that string of data transfered).  But a larger data download of, say, a movie file, would have a dramatically different pattern and the protocol overhead would be quite small in comparison to the data transferred.  Protocol overhead is not some "tax" it is inherent in how the Internet works.  

 



Thank you for your comments. They are much appreciated and contribute significantly to the discussion. I would like to comment on your last item.

 

I think you may have misinterpreted what I was trying to say, or maybe I did not say it well. In any case let me try again. The meter shows all traffic, both high-overhead and low-overhead, across the modem. So say for example I am a huge Internet TV and online movie user. Also assume these activities have an overhead rate of 20% of the traffic associated with these activities. Now lets assume you are a heavy FTP user and the overhead associated with that activity is 10%.

 

You and I both access or send a total of 1GB of actual data content (payload) through our resprctive connections. My meter will show, ignoring rounding for this discussion, 1.2GB and your meter will show 1.1GB yet we both only accessed or sent exactly the same amount of content.

 

Even worse, using your text example, your meter shows 1.1GB based on your FTP activity but your neighbor who only uses the Internet for texting sent or received 1GB of text messages has a meter indicating 2GB transferred due to the high-overhead of texting.

 

The report does indicate a good job with respect to the testing performed and it reports on the accuracy of the meter. I still think the process and procedures were inadequate to address the larger scheme of traffic monitoring, measurement, and reporting. It does not mater how accurate a meter is if it is not capturing and reporting the correct data. Once again this is only my opinion. I have not had the benefit of conducting months of testing and analysis as did the authors of the report.

 

Message Edited by FishMan on 12-06-2009 08:33 AM
Diamond Problem Solver

Re: Data Usage Meter Launched

Thanks for your clarifications and your usual candidness as well Jason. 
Official Employee

Re: Data Usage Meter Launched

Replies in blue

I have three questions
1) What is the purpose of this meter?
 
To enable customers to keep track of their Internet data usage.
 
2) Was the current limit set to curb normal users or to provide an absurdly high limit for abusive users? 
 
The current limit is 250GB.  Please refer to our FAQs on the 250GB limit for more information.
 
3) To highlight Fishman's point, if it's tied to IP address, is Comcast going to reset the meter associated with an IP address when it's reassigned to someone else? Many of us have been falsely accused of spamming and other nefarious activity based on a reassigned IP address.What has Comcast done to ensure we aren't accused of having high data usage? 
 
It is not tied to IP address.
 
What I've heard from Comcast (maybe not what they intended) is that ordinary users won't come close to the limit.
 
Correct.
 
Having this meter can show users that they really do have plenty of room to explore other varieties of data usage.
 
Correct.
 
The limit was set in a time when people weren't downloading high-definition movies as a matter of course. If we see lots of users bumping into the limits, maybe we'll need to lobby for a higher limit. Or maybe it will be a non-issue. I have a feeling that Comcast doesn't really know what a realistic unobtrusive limit is.
 
The limit was established at a time when HD movie downloads were in fact being observed.  The limit is high and fair.  Compared to any other ISP with a limit that I am aware of in North America, 250GB is the highest, and in many cases by a long-run.
 


 

Message Edited by jlivingood on 12-06-2009 09:41 AM
JL
Internet Services
Official Employee

Re: Data Usage Meter Launched

Replies in blue.
 

The report explained a lot but I think it also raises questions.

 

1. The report describes 6.2% measured overhead as a result of the testing. This overhead is added to the meter. Testing consisted of transferring files using FTP which is a low overhead protocol. Why didn't the testers use a higher overhead protocol such as HTTP, which would be more representative of a typical Comcast users experience?

HTTP was not a suitable protocol for us to generate a huge volume of data transfer in a very short period of time.  Also, the point of the exercise was to try to minimize and control for the protocol overhead to the greatest extent possible so that we could more accurately gauge what the PC saw and then what each part of the measurement system recorded.  Thus, the goal was to measure measurement accuracy, and not to simulate the usage patterns of an "average user."  (Then we'd rathole on what average usage patterns are, etc., among other things.)  The other thing that is important to bear in mind is that many other protocols, and in particular HTTP, are highly variable in the overhead observed.  And so I didn't want to turn months of testing into years of testing.  If people have concerned over protocol overhead, I recommend participating at the IETF to make future protocols more data efficient (which often comes at the expense of functionality and extensibility).  ;-) 

 

2. The report goes on to identify several ways and devices that use the Internet that may be unknown to a user. The number of devices and their respective Internet usage will grow in time. There is no discussion in the report of how the bandwidth restrictions and monitoring will grow to accommodate growing use.

 

The NetForecast report is a study of the measurement system and it's accuracy.  It was not intended to address or examine Comcast policy changes, such as changing data usage policies.  However, the current system of using IPDR to collect usage statistics will continue into the future and will scale well, and it is built into DOCSIS 3.0.  I do not anticipate massive changes to the system, particularly as I am in the process of deploying it nationally after an extended development and testing period.  :-)

 

3. It is my understanding from the report that traffic tracking across the modem is captured and uniquely identified by the modem's ip address. The case where an ip address changes due to system maintenance, changed device connected to the modem, or other situation causing DHCP to assign a new ip addess is not addressed. What happens to the meter data when a new ip address is assigned? What happens to a user's account if the newly assigned ip address was one that during the month was previously assigned to a user who exceeded the 250GB limit? We've seen issues where ip addresses that were previously blocked were reassigned to subscribers.

 

IPDR is used to collect data usage statistics.  IPDR uses the modem's MAC address, and not the IP address.  As you indicate, the IP may change frequently, whereas the device MAC address does not change unless the device itself is changed. 

 

The report was good to read but it was incomplete in my opinion. I don't think the testing process and procedures adequately represented a typical Comcast user's traffic.

 

As indicated above, that was not a goal of the report.

 

It seems to me that traffic overhead should be a cost to Comcast of doing the business of providing Internet service. Measuring overhead and including this number in the usage meter is akin to FedEx charging $28 for a package then adding $1 for gas plus $1.50 for rent, plus ... Comcast IMHO should charge for the internet connection and not the overhead separately.

 

It is technically impossible in actual practice, and at reasonable expense, to separate out the protocol overhead from the actual data transfer.  The only way I could see to do it, which I would be strongly against, would be to deploy a massive (and I mean MASSIVE) Deep Packet Inspection system to examine the contents of all user traffic and perform content analysis and somehow separate out protocol-related traffic from data transfer.  This would involve a deep study of every protocol in existence and would be a moving target since protocol overhead, even for the same protocol, is HIGHLY variable and not fixed.  I simply don't see how such a system would be technically practicable or desirable.  Furthermore, it is not the state of the art to extract out protocol overhead when calculating data transfer.  It is understood that data transfer is inclusive of protocol communications, and this is how all data transfer measurement that I am aware of functions. 

 

Wait until they decide that Netflix users need to pay 10% overhead, HTTP users should pay 12% overhead, Internet TV viewers should pay 25% overhead, etc. Oh wait, that is exactly what this metering system does. If you are a high overhead protocol user you will reach the limit faster than a low overhead protocol user.

 

This is inaccurate and you are reading far too much into our statements about protocol overhead.  Without the exchange of protocol communications, not other communications can occur.  Protocol overhead is extraordinarily variable.  A very short communication, such an an IM, may have a lot of overhead relatively to the small payload of a message saying "Are you there?" (in such case, the protocol overhead would likely exceed that string of data transfered).  But a larger data download of, say, a movie file, would have a dramatically different pattern and the protocol overhead would be quite small in comparison to the data transferred.  Protocol overhead is not some "tax" it is inherent in how the Internet works.  

 


JL
Internet Services
Silver Problem Solver

Re: Data Usage Meter Launched


bethkatz wrote:
Fishman, good points and good suggestions for what they should have included. But maybe they didn't need to. Maybe the limit should go up another 25% to include the overhead of other media.
 
I have three questions
1) What is the purpose of this meter?
 
So customers can tell when they're nearing the 250 GB/mo bandwidth cap. 
 
2) Was the current limit set to curb normal users or to provide an absurdly high limit for abusive users? 
 
The latter. See the FAQ. 
 
3) To highlight Fishman's point, if it's tied to IP address, is Comcast going to reset the meter associated with an IP address when it's reassigned to someone else? Many of us have been falsely accused of spamming and other nefarious activity based on a reassigned IP address.What has Comcast done to ensure we aren't accused of having high data usage? 
 
I belive it's tied to the modem, not the IP. 
 
What I've heard from Comcast (maybe not what they intended) is that ordinary users won't come close to the limit. Having this meter can show users that they really do have plenty of room to explore other varieties of data usage.
 
The limit was set in a time when people weren't downloading high-definition movies as a matter of course. If we see lots of users bumping into the limits, maybe we'll need to lobby for a higher limit. Or maybe it will be a non-issue. I have a feeling that Comcast doesn't really know what a realistic unobtrusive limit is.
 
As ordinary use of the Internet grows, I full expect that Comcast will adjust their limit. Remember, to be an excessive user, you have to be above the cap and one of the heaviest users. If 50% of users are above 250 GB, simply going above that level obviously won't make you one of the heaviest users. For the limit to be useful, they'll need to adjust it so that it corresponds to the truly excessive users.
 


 

Silver Problem Solver

Re: Data Usage Meter Launched

EG, I hope that you are right. But I was accused of having an infected machine knowing darn well I was falsely accused. They cut off port 25 for outgoing email with no warning. And we've seen it happen several times with other customers.

 

Jason, could you please let us know that Comcast is sure that the usage meter is tied to each customer's modem and not to the changing IP number?

 

Do we have any customer feedback from people who have the data usage meter available now? 

Diamond Problem Solver

Re: Data Usage Meter Launched


FishMan wrote:

 

EG, thanks for the clarification on point 3.


I could be wrong about that Fishy. As I said to Beth, a confirm by a CC employee would be nice.

Diamond Problem Solver

Re: Data Usage Meter Launched


bethkatz wrote:

 

3) To highlight Fishman's point, if it's tied to IP address, is Comcast going to reset the meter associated with an IP address when it's reassigned to someone else?  


Beth. See my last comment. I could be wrong. A authoritative confirmation by a CC employee in the know would be nice here.

Message Edited by EG on 12-05-2009 07:48 PM
Bronze Problem Solver

Re: Data Usage Meter Launched


EG wrote:

FishMan wrote:

The report explained a lot but I think it also raises questions.

 

1. The report describes 6.2% measured overhead as a result of the testing. This overhead is added to the meter. Testing consisted of transferring files using FTP which is a low overhead protocol. Why didn't the testers use a higher overhead protocol such as HTTP, which would be more representative of a typical Comcast users experience?

 

Way Agree !

 

2. The report goes on to identify several ways and devices that use the Internet that may be unknown to a user. The number of devices and their respective Internet usage will grow in time. There is no discussion in the report of how the bandwidth restrictions and monitoring will grow to accommodate growing use.

 

Agree here too.

 

3. It is my understanding from the report that traffic tracking across the modem is captured and uniquely identified by the modem's ip address. The case where an ip address changes due to system maintenance, changed device connected to the modem, or other situation causing DHCP to assign a new ip addess is not addressed. What happens to the meter data when a new ip address is assigned? What happens to a user's account if the newly assigned ip address was one that during the month was previously assigned to a user who exceeded the 250GB limit? We've seen issues where ip addresses that were previously blocked were reassigned to subscribers.

 

I believe that it is the Comcast private modem management IP address that they use here, and not the subscriber's public (WAN) IP addy that is assigned by CC's DHCP server to one's connection.

 


 

Message Edited by EG on 12-05-2009 07:12 PM

EG, thanks for the clarification on point 3. Smiley Happy

Message Edited by FishMan on 12-05-2009 04:43 PM