Forum Announcement

Welcome to the new CNET Forums! Please don't panic. You are not in the Twilight Zone, you are experiencing the new CNET forums platform! Please click here to read the details. Thanks!!

Internet Services Providers

General discussion

Where is all my bandwidth going? And how do I tell?

by Lee Koo (ADMIN) CNET staff/forum admin / December 13, 2013 7:13 AM PST
Question:

Where is all my bandwidth going? And how do I tell?


I don't have fiber to the home, so I am limited to my cable Internet connection. I'm currently paying for the fastest residential Internet my cable provider offers. Using Speedtest.net, I routinely get 20Mbps down and 4Mbps up. Advertised speeds exactly. We have TWC/Road Runner in the San Antonio area.

We have wired Gigabit Ethernet in the house.
iMac (Mavericks) (on all the time)
Sony PC (Windows 8.1) (on all the time)
HP LaserJet P2055dn (on all the time)
Two DirecTV HD DVRs (standby or on all the time)
Apple TV
Xbox 360

Two "Smart" TVs
Pioneer AV Receiver (AirPlay)
Two Sony BluRay players (one is a combo AV receiver)
Apple Gigabit Router (on all the time)
File Transporter (on all the time)

All of these are wired. Occasionally we'll connect a laptop, usually a MacBook Air or Pro, to the Ethernet instead of just connecting wirelessly. We have two iPhones and an iPad connected wirelessly to the network when they are home. We rarely connect the laptops to the network wirelessly -- usually just pick up the iPad. Then of course there are the apps (that I know about) that are running on the desktops. Dropbox, SkyDrive, FileTransporter, and iTunes. We hope to eliminate (or consolidate) and only use SkyDrive, once our workplace adopts it/allows it in addition to Dropbox. Until then, we are using all three Dropbox-style apps.

I think the performance issues we are having are community-related, and not necessarily our own doing. I have taken the time and effort to go around and unhook each device and retest. I really thought DirecTV was the hog, but disconnecting both simultaneously did nothing to the network speed. At various times tonight, after 6 p.m., the network seems to really slow down. Netflix, for example, will often go from a nice near-HD resolution to really bad, pixelated/JPEG artifact-looking quality to keep up. A DirecTV on demand that comes through the Internet might take 30 minutes before we get the "ready to watch" prompt. Off the satellite, voila, instant movie.

Even during this time, Speedtest shows advertised speeds. Hmmm. Download something and it sure seems slow.

Is it me, or my cable provider, reaching capacity?

Thanks.

-Submitted by: Jack P.
Post a reply
Discussion is locked
You are posting a reply to: Where is all my bandwidth going? And how do I tell?
The posting of advertisements, profanity, or personal attacks is prohibited. Please refer to our CNET Forums policies for details. All submitted content is subject to our Terms of Use.
Track this discussion and email me when there are updates

If you're asking for technical help, please be sure to include all your system info, including operating system, model number, and any other specifics related to the problem. Also please exercise your best judgment when posting in the forums--revealing personal information such as your e-mail address, telephone number, and address is not recommended.

You are reporting the following post: Where is all my bandwidth going? And how do I tell?
This post has been flagged and will be reviewed by our staff. Thank you for helping us maintain CNET's great community.
Sorry, there was a problem flagging this post. Please try again now or at a later time.
If you believe this post is offensive or violates the CNET Forums' Usage policies, you can report it below (this will not automatically remove the post). Once reported, our moderators will be notified and the post will be reviewed.
Collapse -
Shrinking bandwidth
by flrhcarr / December 13, 2013 9:51 AM PST

Aside from cable limiting the amounts we can use. . . There are also many subscribers on the line, or could be. It's nice if you have a dedicated line to yourself, but cable can sell that line quite a few times. This is obvious in communities, & apartment complexes. When I lived in one a few years ago, I could tell when someone was moving in, because the strength dropped.
I myself have a horrible time getting things to load using wifi, but most of that, it because there are others in the house, using it.
When connected with a wire, you still have to combat that cable does limit your amount. They even released something a few years ago saying that they would. Something like those who use it less, will get the most bandwidth. Which you can test at the 1st of every month.

Another thing is how you are wired? USB isn't great, but it is consistent. I actually switched to it for a couple years, because the router just didn't "like" the cat cable connection.

That being said; Check connections. See where your feed it going? See if maybe your computer isn't pulling in as much as it could. Are your pages (that you visit) over loaded? A newer version of your browser may not help, all of the time, but it may some of the time.

Then get in to what your computers are accessing? Updates are the worst, & Adobe is the king of bandwidth killers!

Seeing that you've already unplugged everything, chased all the dust off the connections & reset your cable boxes, I'd go with limited usage, if everything else adds up. Good luck.

Collapse -
I agree
by kanehi / December 20, 2013 9:45 PM PST
In reply to: Shrinking bandwidth

Cable broadband is a shared line and when other subscribers are logging in the speed will decrease. This is why the speed decreases at certain peak times of day specially when people are just coming home. Also make sure your system's firewall are in place since it's a shared line people can access your system.

Collapse -
Also...
by JCitizen / January 7, 2014 6:20 AM PST
In reply to: Shrinking bandwidth

Since he uses wired networking primarily, readers here might be aware of collisions of the data packets that can cause bottlenecks and slow downs. Sometimes folks who have an extensive network in the home may be using hubs instead of switches. These can cause a lot of packet collisions, and force other devices to do the QOS work, which can slow the TX and receiving devices too. Hubs should be avoided in these architectures, and switches would be better - even a hub with auto-sensing dual speed, and/or is stackable, and/or at least has SNMP support can improve the situation. A switch is better.

This is not the condition here, because Jack P. did a test removing all competing devices, but it can help to be aware of these things. However, even with one device and a modem, the cable between them should have the fewest connections possible, or use the highest speed as the "backbone", so in the daisy chain, you want your main cable connected to the 100mps, or 1000mps ports in your switches/routers. If you rely on a NAS to provide HDTV streaming for an entertainment display or similar device, this can make these factors especially important, because the ISP connection has nothing to do with the interior networking speed you need with in home streaming. That is unless there is a DRM issue, but that is a whole other discussion.

Also length of cable and the rating of the cable can affect you if you are doing a lot of HD streaming. This can affect the network behind the router(LAN) because some folks are lucky enough to have an ISP with gigabyte service, and a lot of older networks only run 10/100mbs or use G wireless or slower. I'm only making these comments because he describes a fairly complex home network, the more connections you have, and the type of service you expect to use can bring up newer specifications as requirements. I'm sure subsequent comments here will cover this better than I am.

Collapse -
No, No, and no
by mrgale--2008 / December 13, 2013 10:27 AM PST

What you are having is a variation of the "last mile" problem. Your speed is limited by the lesser of the your ISP speed and the speed of each segments between you and your destination. Testing with my ISP, Comcast cable (105 Down/20 Up), I get the following speeds using speedtest.net - 68/20, 67/20, 121/23, 40/23, 32/15, and 6/7. The last is to Alaska. Put another way, your speed is governed by the slowest link in the chain, which is generally not your ISP

Collapse -
Traceroute
by phlashadelic / December 20, 2013 10:41 AM PST
In reply to: No, No, and no

The traceroute command line utility can also shed some light on just exactly where the slowdown lies.

Sometimes there are very many routers between the server at the service where the video stream is coming from and your home device that is displaying the video. Any one of these can cause a problem with slowing down - especially if your stream has to pass through one of the congested network access points (NAP) where traffic is handed off between high level Internet service providers.

I have provided the results of a traceroute from my house to the host at www.samsung.net
I figured it would be a rather lengthy route in order to help illustrate my point.

C:\Users\Username>tracert www.samsung.net

Tracing route to www.samsung.net [203.254.227.78]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms LocalNetWorkName [ <ipaddress><span id="INSERTION_MARKER">IPaddress ]
2 17 ms 15 ms 10 ms router name <routername.domain.com>[<ipaddress><span id="INSERTION_MARKER"> IP address ]
3 79 ms 102 ms 71 ms router name <routername.domain.com>[ IP address <ipaddress>]
4 23 ms 15 ms 17 ms router name <routername.domain.com>[ IP address <ipaddress>]
5 35 ms 33 ms 44 ms router name <routername.domain.com>[ IP address<ipaddress>]
6 52 ms 61 ms 53 ms router name <routername.domain.com>[ IP address <ipaddress>]
7 60 ms 51 ms 62 ms xe-7-2-0.edge4.Chicago3.Level3.net [4.53.98.41]
8 91 ms 56 ms 69 ms ae-2-52.edge3.Chicago3.Level3.net [4.69.138.168]
9 69 ms 68 ms 66 ms GBLX-level3-4x10G.Chicago.Level3.net [4.68.62.254]
10 86 ms 82 ms 91 ms 146.82.32.22
11 224 ms 228 ms 228 ms u13.ppp66.samsung.co.kr [157.197.66.13]
12 224 ms 227 ms 227 ms u86.ppp66.samsung.co.kr [157.197.66.86]
13 222 ms 218 ms 211 ms u126.ppp82.samsung.co.kr [157.197.82.126]
14 224 ms 228 ms 227 ms u2.ppp83.samsung.co.kr [157.197.83.2]
15 * * * Request timed out.
16 237 ms 231 ms 225 ms 203.254.226.69
17 232 ms 232 ms 233 ms 203.254.227.78

Trace complete.

C:\Users\Username>
</ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress>
<ipaddress><routername.domain.com><ipaddress><routername.domain.com><ipaddress><routername.domain.com><ipaddress><routername.domain.com><ipaddress><routername.domain.com><ipaddress>
</ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress></routername.domain.com></ipaddress>
I have blocked out information up until the handoff to the large network provider Level3 at the Chicago NAP

As you can see, speed tests may be completely meaningless if your traffic has to encounter heavy traffic OR needs to travel halfway around the world from its source.

I hope this helps.

Collapse -
Where is bandwidth going?
by tcritchley07 / December 20, 2013 4:29 PM PST
In reply to: No, No, and no

mrgale reflects my thoughts exactly. if you go on a multi-stage journey, last stage by jet plane (your local superfast last leg from exchange to your PC), you think the whole trip should be fast. If stage 1 is by car, stage 2 by train and stage 3 on foot, it doesn't matter how fast the jet travels, it's still going to take longer than you thought. There's also the response time of the web site with time going to 'waiting for web site XXXX'. It would be useful to be able to investigate the 'hops' taken by your traffic in different situations ('hops' are different routes thru a network). Not techie enough to tell you how to do this.
Throttling down at local level has also been mooted and I've heard this too as a possible reason.(I've just seen the note on TRACEROUTE below my reply!!!!!). That's the thing I was searching for.
Terry Tortoise

Collapse -
where the flix hit the net
by dangoodale / December 13, 2013 11:14 AM PST

Jack,

I have a very similar issue. I was upset with my provider thinking they were the cause of my NetFlix blues but, it turned out not to be the case, in my case. First I ran a speed test which showed that my download speed was still VERY fast. But, I'm the suspicious type and if I ran a cable company and was oversubscribing my clients I MIGHT think about giving a bit more bandwidth to clients trying to perform speed tests. Sneaky, but doable.

THEN I noticed that while the image on my TV screen was all pixelated and nasty, the picture on my COMPUTER screen was just fine watching the same movie at the same time. That made me suspect my Netflix source, my Roku box (which was running wireless, exactly 11 feet away from my wireless N router). I ran a wired connection to my Roku and, for the most part, my problem went away.

My point isn't that a Roku or even wireless is the culprit. My point is it can take a lot of investigation to figure out what is causing the problem. Since you, for the most part, are running everything wired, you should probably consider everything in the path between your router and your TV and even eliminate a lot of it altogether for a sanity check.

I'd try a known good patch cable plugged directly into my router, with everything else unplugged from your network. If that seems to resolve the issue, work backwards from there. You could have a bad patch cable to the tv experiencing the problem, or a bad network card in one of your devices that is causing difficulties for your other devices, or one device could be running at half duplex while all the rest of them are running at full duplex which would cause packet collisions. (So, check the connection settings on ALL your hardwired devices.)

You might also want to check out your router and see if it can keep records of how much bandwidth you're using. Mine can (and was no help figuring this out). Wink

Good luck!

Dan Goodale

Collapse -
The only option left is...
by aswnc / December 13, 2013 11:33 AM PST

Bandwidth throttling. Every week I have a good answer to these and never get to them in time. The week I finally do, it's a tough one to answer (or at least to prove). Based on what you've said - primarily that the speedtest shows full speed, it almost has to be bandwidth throttling by your ISP. Speed tests generally use a relatively small file (about 1MB or so) for the test so people with slow connections don't have to wait so long. The ISPs, in some cases, throttle once the usage hits a certain amount. That way, the casual user isn't affected and doesn't complain. When you're watching a video, it starts ok but when you go over a certain amount, they may be throttling you back.

The first thing to do is simply ask your ISP. Maybe they'll be straight up with you. Another option, if you live in a good area, is to borrow a friends 4G hotspot and test. (You may need to buy a router with wireless Internet connection - usually a USB port - to mimic your current network.) You could also possibly talk to a 4G provider to see if they would let you try it - with a slight indication you may switch if it works. A final option is to find someone else with the same provider and speeds as you and try to mimic your situation in their home. Once you've been able to compare your current setup with another, you can tell for sure if it's your devices or the ISP. Hopefully the ISP would be truthful up front and save you the time.

There is one other possibility. Beyond actual speeds, latency is another concern. Speed tests may show latency - usually just one test with the ping. A slow ping could indicate a hesitation for connectivity or handshakes, followed by normal/fast speeds. The streaming movie service will account for intermittent connections and provide lower res to avoid buffering pauses. If that's the case, you could look at anything from bad cables, to bad equipment, to apps/viruses/spam, to the ISP. I'd start with the throttling bet first. Good luck.

Collapse -
What is your main Ethernet distribution?
by SD_DS / December 13, 2013 2:32 PM PST

I see 13 items on the wired network, but no mention of what's really driving them. There should be a cable modem, a gateway router, and large switch (or several switches) with a large number of ports feeding your Ethernet connections (the router is often combined with a 4-port (or larger) switch). The Apple router only has a couple of switched ports, but to use them, it would have to be configured in a pass-thru mode, since it apparently isn't your gateway router. When you find the router & switch(es), you might be able to use their tools to see what's going on with your network. I have a small Gigabit wireless router (with 4-port wired switch built-in) that allows me to optimize performance for video streams; perhaps your system has something similar.

Note, too, that your ISP speeds are far below your Gigabit Ethernet rates. Are you sure that all ports are at Gigabit speeds? If you have some switches that are the usual 100 Mbps, you won't notice a speed difference on a speed test, but when you are trying to share the bandwidth with many other devices, they won't switch off the network as fast and could appear to be a bandwidth problem.

I agree that some of your problems could be the community issues. I'm also skeptical of TWC's treatment of your Netflix and DirectTV. TWC wants to sell you the movies on demand and DirectTV and Netflix are their competition. Do you think they will be completely net neutral sending data from their servers? I would guess that the internet version (for play on a tablet or PC) is smaller in bandwidth and different in format than something for a full 1080P video screen. It might also be easier for Netflix/DirectTV to send those versions from a different server; that would make it very easy for TWC to bandwidth limit the direct-to-TV version coming from the main server. This would be difficult to prove, although if you tracked where the streams come from you might be able to make the case to Netflix or DirectTV. I'm sure they'd help you get your full bandwidth back for their products.

Also, many programs that auto-update do so constantly -- every few minutes or even more often. With so many devices on the network, some of your slowing traffic may just be unnecessary update requests or data uploads to the various home servers. I turned off Windows auto-update when I saw it pinging Redmond, WA every few minutes at 2 AM on a Sat night. Microsoft isn't going to update at 2 AM on a Sat night, so this is just wasted bandwidth. I found that having Windows notify me that there were updates eliminated most of the active pinging, and all I have to do is enable the update when notified. Many other auto updates are as annoying as Microsoft, so I switched them to notify as well. Your "smart" TVs and other networked A/V items may also have auto-update traffic that can be turned off, too. Of course, turning off the apps, at least for a while as a test, would be a good idea.

Another bandwidth sucking item are cookies on your systems. Many cookies are used to track everything that is looked at on the web. When you click on something, not only does that request get sent, but possibly dozens of tracking cookies reporting back to home base about what you just did. Again, wasted bandwidth! I set my browsers to clear all cookies after I close the browser; I use exceptions on the few sites I actually want to log into again in the future. This eliminates all of those unwanted tracking data packets being transmitted out of your systems -- improving your apparent bandwidth.

Collapse -
One router, three switches
by El_vis / December 14, 2013 12:38 AM PST

The Apple router is connected to the cable modem. The ethernet ports on this router go to the primary computers (iMac and Sony PCs) and to a Netgear ProSafe gigabit switch. This switch is connected to two other switches (Netgear consumer variety) at each of the two TVs plus the printer, file transporter, and to the ADT security system I failed to mention in my original post.

All the entertainment stuff at each TV is connected to these switches. So even though the backbone/infrastructure is all gigabit, most of those components at each TV are only 100mb fast ethernet. I think the new AppleTV might be gigabit, I'd have to check it's specs. I don't really care about that anyway, because I think the infrastructure is way faster that the cable service. I'm getting gigabit speeds between computers on the LAN, which is what I wanted. I never expected to get anything better than cable modem speeds for anything coming from or going out the cable modem.

I called TWC and asked about throttling. No official response yet although the person I talked to admitted that peak times can degrade performance in my area.

Collapse -
Thanks for the update! Data rates can be misleading
by SD_DS / December 15, 2013 7:24 PM PST

There are many Netgear ProSafe gigabit switches, so this is a generic reply. Looking at the Netgear website, several of those switches offer QoS (Quality of Service) processing, allowing the user to tailor which data stream, in your case video, gets priority. This can be important when there are many data streams using a single port -- and that's what your switch does when it talks to the Apple router.

While some of your backbone operates at gigabit speeds, the slower parts are in the 100Mbit range. While this is 5x the cable speed, what happens when 5 things want to talk at once? Ignoring the way Ethernet operates, all of a sudden your data I/O has max'd out your cable. Unfortunately, Ethernet isn't that efficient. Your switch will have to negotiate with the devices and determine how to best send and receive the data. That takes time -- 10 times longer with 100Mbit vs. gigabit. When upgraded my modem and router to both use gigabit speeds (no change in ISP), I saw a nice speed increase simply because all the networking handshakes were now 10x faster than what they were before.

If your switch offers QoS, you should take advantage of it and set it so your streaming devices get priority. If it doesn't, you should seriously consider upgrading to one that does. There are some switches that offer multiple output ports; you might consider one of those and move your computers from the router to the switch to let the switch better utilize the gigabit ports. Unfortunately, Apple doesn't mention QoS for the router; Netgear does make routers with QoS capability (that's what I have).

Another possible alternative would be to move the computers to the switch and put your two main streaming devices on the router gigabit ports. This would have them only competing with each other and the switch port, so you'd likely see an improvement in your video vs the extra layer with the switch. Also, if you want to print from the computers or have the computers talk to the File Transporter, they wouldn't have to compete with your video stream as they do today.

Network topography and configuration can make a difference!

Collapse -
What i think might be the problem
by elggy / December 20, 2013 8:52 PM PST

since I see your using mostly apple products for your network the netgear switches might very well be the problem.how many switches in total do you have are they all working at the same speed's? IE all 10/100/1000 or 100/1000, 10/100 find that out like someone earlier said the slowest link in your network can be your problem especially if you do not have every thing set to run at full duplex speed's. so if you got a gig coming out of your router going to either a gig or a 100 switch things will slow down very fast like all because your 100 switch can not keep up with the gig"1000" switch its to fast for the traffic your trying to us. I run mostly linksys routers and i have a cisco Switch i always make sure that every thing is capable of working 10/100/1000 speeds that way it can manage the speeds better by its like able to slow down every thing to the slowest speed on the network at the time of the fastest and slowest are being used. think of it this way ok if you fill two cooking pots identical pots remind you one 1/2 way up and the other 3/4 full put them on the stove which is going to heat up faster? if i am not mistaken its the one 1/2 full all because it doesn't have as much water to heat up. basically its simple you got one or two connections on the network not at the same speeds or something in the middle that is not cable of letting everything run at the same speed your killing your own network by your own hands for not having every thing set and or connected together right

Collapse -
Cookies don't impact bandwidth
by TyCahill / December 20, 2013 8:16 AM PST

I wouldn't consider cookies to be "bandwidth sucking." Cookies have a maximum size of 4 KB. Most web pages today (with their images, CSS, JavaScript, etc.) are several hundred KB in size. Many websites won't work correctly if you completely disable cookies, and clearing cookies when you close your browser won't help your performance while you're online, because the cookies have already been sent to you while you were online. Deleting cookies definitely helps with privacy, but not bandwidth.

Collapse -
It isn't the cookies, it's their outgoing messages
by SD_DS / December 20, 2013 9:02 PM PST

It isn't the transmission of the cookies to your computer that's the problem, it's all of the messages they send out while monitoring your actions. If you haven't cleared out your cookies on a regular basis, you could have several dozen all trying to report back to their masters what you just clicked on, or even just hovered over, in a page. They each are sending data to different computers; each one of those messages requires a new data packet set for each server they send the data to. Since apparently QoS isn't utilized, any streaming data handshakes can be waiting in back of the line of data to be transmitted. No handshake = no stream. With several dozen messages queued up, or several groups of several dozen messages (multiple mouse movements/clicks), the handshake delays can appear to be a bandwidth problem, while, in fact, it's a cookie problem.

By regularly clearing third party tracking cookies, you will minimized the number of outgoing messages and ensure that your data streaming handshakes are processed as fast as possible, especially if you aren't using QoS to help your streaming.

Collapse -
Cookies are passive
by MightyDrakeC / December 21, 2013 3:25 PM PST

Cookies are *not* little programs nor snippets that actively run while you use your browser. Cookies don't "...report back to their masters..." on their own. They just sit there passively until a particular server queries your browser for a particular cookie. Then, your browser will send the contents of that cookie back to the server.

If it's the first query from that site since you last cleared cookies, then the query will report "Not found" and the server will then send the cookie to your browser, which will then save it, to be queried for all subsequent requests. Those subsequent requests will be during that same browsing session.

What all that means is, except for one query per cookie, the network traffic will be nearly identical whether you clear cookies often or not.

I think what you have in mind is, some sites have Javascript code which will change the behavior *of that page* based on information in the cookies. But, Javascript runs locally, so that won't affect bandwidth. And, the effects will be limited to that page. Unless you have something like a malware/spyware toolbar installed.

Cookies are data. They are not code. And, they tend to be very small. It is only in pathological cases (meaning, very poorly written website code) where they can have a significant effect on bandwidth or latency.

Drake

Collapse -
Not all cookies are the same!
by SD_DS / December 29, 2013 9:30 PM PST
In reply to: Cookies are passive

Yes, the cookies are just data, however, having them on your computer allows their servers to repeatedly query your system for changes/updates to anything they've stored in the cookies. If that includes your mouse position or webpage, the server will continue to ping your system relentlessly for updates. Browsers have even tried to limit this by adding the "Don't track me" option that is often skirted by the tracking folks. "Pathological" is probably one of the kinder terms used for the tracking cookie designers.

The original question was about bandwidth and video streaming. While a normal cookie, as you described, is rarely used -- like an encoded word that helps your bank verify this was the computer you last used to logon with -- tracking cookies are used for tracking your every move while on the internet. The increased network activity will result in less bandwidth available for other data -- like video streaming. The fewer cookies you have on your system, the fewer servers your system will have to respond to and the lower the bandwidth used while responding to those servers.

Collapse -
Cookies really are passive. I promise
by MightyDrakeC / December 30, 2013 1:42 AM PST

I'm afraid you have misunderstood what tracking cookies are and how they work.

Cookies are data. They really do just sit there. They initiate nothing on their own.

Another attribute of cookies is that they are tied to a particular website. CNET can have more than one cookie. For example. A global CNET user ID cookie, and a CNET forum cookie, that the developers of the forum software use to keep track of forum-related info, etc.

But, there is no mechanism where CNET can ask for cookies of other websites. CNET cannot ask for your Amazon cookies. CNET does not query the browser for, "The CNET global user ID cookie." Instead, CNET queries the browser for "The cookie called global user ID" and the browser looks for a cookie by that name among its collection of CNET-specific cookies.

But, there are what are called third-party cookies. The way a webpage is put together, some sections of a webpage can come from a different server than is in the address bar. One of the most common uses is for a webpage to set aside a banner or box of a certain size, and that chunk is filled by calling an advertising server.

The code used to build that block of the webpage can also store their own cookies. Those cookies are also strictly tied to the web server of that code. Therefore, the advertising server cannot query CNET cookies. They can only query their own cookies.

But, that has its own advantages. If the advertising service is being used by multiple websites, then the cookies that the advertiser is querying will be identical when the user is on multiple sites. So, the advertisers give the browser their own user ID cookie. With that, they know that a particular browser on a particular computer has been to Sites A, B, and C, which all use their service for ads. But, they will not know if that browser has been to Sites X, Y, or Z, if those sites do not use their service.

With that mechanism, advertisers can learn some of the websites you have been to. And, depending on how the websites interact with the advertisers code, they can know which pages on those websites you've been to. And, depending on the details of the relationship between the website and the advertiser, the website might provide other information, such as the user's name and address, search phrases the user used within that website, or other more detailed information. That last would probably not be handled with cookies. They would likely have some other programmatic conduit between the website's server and the advertiser's server.

But none of those are "...cookies reporting your mouse movements..." And, certainly, none of those are rogue code snippets where multiple advertisers can watch what you're doing on every single site you visit. Advertisers can only get info from websites where the owners of those websites have chosen to insert the code to call on the advertiser's server. So, except for possibly some porn sites, you're not going to have dozens of advertisers on one page at the same time.

I hope this clears up for you what tracking cookies are capable of and what they're not. Cookies are not malware. There are privacy considerations with any web cookies. (Albeit, I'm pretty laissez faire about that sort of thing.) But, they are nowhere near nasty as you have thought they were.

Drake

Collapse -
Clarification
by MightyDrakeC / December 30, 2013 1:58 AM PST

I just reread what you wrote. You seem to be under the impression that having a cookie on your computer somehow gives the issuing server some mechanism to query that cookie at any time. That's not how it works.

Cookies are queried only when your browser initiates a connection to a server. Your browser does that based on the code of the page it's currently looking at. In the case of third-party cookies, a particular webpage can kick off queries to multiple servers. But, only those servers will respond and query for their cookies.

None of the other dozens or hundreds servers that have cookies currently sitting on your hard drive have any way to query their cookies. They can only query them when you navigate to a webpage that initiates a connection to those servers. The servers cannot kick off a query on their own.

So, no, having cookies does not result in server queries at any random time that you're using your computer, nor when you're actively browsing other sites. Only when you hit sites that have pages that specifically reference those servers.

Drake

Collapse -
I would still concider..
by JCitizen / January 7, 2014 8:55 AM PST
In reply to: Clarification

what the mission of the servers are from the various sites. When you look at what NoScript is blocking when you 1st visit a site, it becomes obvious that some sites just want to get deep in your shorts, and gather/report way too much information, or use bad web coding, or many other forms of bad commerce related to cookies. Many times they aren't malicious per se, but just bad practice. I know I definitely see the performance difference in page loads, when I have 3rd party or maybe even all cookies blocked. However, since many sites require at least a minimum cookie tolerance, one cannot always do that. I also must mention that cookies can prevent repetitive popups or other promotional server behaviors when they are used wisely by web sites to help the server know you have completed an action or seen a page already and gained those experiences before.

It still seems to me that when you get too much of these special text files in there, and you do hit a site that uses bad practices, a good cookie cleaning definitely helps, because at least only the cookies placed by that site are doing any interaction with the server, and no tracking information is being blabbed back and forth from the web site. I set my cleaner to keep cookies from sites I like, and this seems to help establish a little faster browser performance too - occasionally - because the cookie is already set, and the site servers have nothing to do generally but read them, as a minimum, and not set a new one, or add anymore than is necessary, new information in them.

I think most people would do better if they used a passive protection utility that blocked some of these badly practiced schemes in the 1st place, which seems to blind over-reaching communication with said poorly instituted services. It does little good to just avoid the site, because there is usually something useful for the client visiting it. What happens when malicious intent occurs is when something that is actually an active malicious malware gets deposited in the temporary files and goes filching around in account files(including cookie data), up to no good at all - having cookies set by these malicious processes can't be good that is for sure, (for instance - If a webserver sets a cookie with a secure attribute from a non-secure connection, the cookie can still be intercepted when it is sent to the user by man-in-the-middle attacks) So sometimes a cookie is just one more thing you don't want in your computer, and especially if malcode get into the app-data files trying to subvert processes without requesting privileges from the operating system. Most of my clients can't believe how much mischief malware can cause without setting off the UAC, or even a good HIPS utility. As far as I'm concerned I don't want ANY of their files in there, and if my passive protection can block cookies, so be it - fortunately some anti-malware are very good at blocking the IP address of known bad servers anyway, so this can also be prevented.

It has never surprised me how much better the performance of a client's web browsing can become after installing a good firewall, passive protection utilities, and host files to block most of these bandwidth hogging servers, and yes, I don't want their damn cookies in there either. Especially NOT zombie or persistent cookies(LSOs). There is definitely more information about you in some of the cookies, that you may only want a single web site to know, but under malware direction, outside minion servers may receive all the information in the cookie folder, and sensitive meta data too! I mean - let's face it - when malware can read and find a credit card number or SSN in your hard drive in fifteen seconds, I got a problem not believing they cannot break into the cookie folder and do pretty much what they want to there also.

I also want to acknowledge that even good sites can become infected and serve up all kinds of trash; so I'm not always blaming a site that originally had good intent, but for web administration security should still be a priority if they are worth doing business in the future.

Collapse -
Cookies are not a bandwidth issue
by MightyDrakeC / January 7, 2014 10:37 PM PST

I'm afraid you still don't have a good grasp on what cookies are and how they're used by the programmers of websites.

First, bandwidth, the subject of this thread. The likelihood that cookies have a noticeable effect on a video stream approaches zero. It just isn't happening. Cookies use a trivial amount of bandwidth.

Second, cookies as a privacy issue. To someone who is hyper-concerned about privacy, the standard cookies and Flash cookies (LSOs) are an issue. That's a whole other topic of discussion. Each person has to determine his own level of comfort on the ability of various companies to catch glimpses into his surfing habits.

And the third thing you brought up, malware. If you have malware on your system then cookies are the least of your worries. Your system is completely compromised, and nothing is safe. But, one thing is probably true. The rogue program is almost certainly completely bypassing the browser cookie system to do its dirty work.

You appear to be conflating privacy issues with security issues, and then trying to cram performance issues in with them. There is some overlap. NoScript in particular can touch on each of them, albeit by blocking Javascript, not cookies. But, for the most part, the three subjects are pretty distinct. And cookies are almost solely a privacy issue. They rarely have any effect on security. And it's nearly impossible for cookies to become a performance issue.

Drake Christensen

Collapse -
All I know is..
by JCitizen / January 8, 2014 6:45 AM PST

When I block all of that crap - my "bandwidth" is fine, and I then have no problems with Netflix or any other service. This is repeat performance for every client I work with and solves the issues. I takes running the gamut to get things done unfortunately.

Collapse -
I feel your pain.
by Aseriesguy / December 13, 2013 3:51 PM PST

We also have cable internet access. We also have numerous devices. Anywhere between 2 and 6 laptops, my son's desktop , 3 tablets, 3 cell phones, a Wii, my son's Xbox360 and a Sony Bluray player we use heavily for Amazon Prime video. I have the latest tech wireless router, a wireless extender and a gigabit switch.
Everything works great 90% of the time but I am convinced half of slow downs are at the server end. Tonight we were streaming video from Amazon Prime as we do most nights. Suddenly, the Sony Bluray repeatedly reported NETWORK ERROR or NETWORK DOWN. The same content was fine on my living room laptop. A reboot of the Bluray got it going twice. I know everything that goes on in the laptop but the Bluray innards are a mystery. I am guessing there is a minor difference in protocol handling on the Bluray.
Just remember that on a cable you are sharing a finite amount of bandwidth with an unknown number of other users in your neighborhood.

Collapse -
saony bluray player is the suspect
by lt x / December 22, 2013 2:19 AM PST
In reply to: I feel your pain.

We have a 5 year old Sony blue ray player. 70% of the time we use it to access Netflix or other movie streaming services, we have pixel images and lockups. We rarely had that problem when we streamed to our laptops, including XP laptops. I suspect that the older Sony blue ray player is the culprit when we use it to stream content on-line. And, Yes, we do download their updates, but that has never solved the problem.

Collapse -
interesting
by elggy / January 9, 2014 1:51 AM PST
In reply to: I feel your pain.

I have had internet and phone modem issue's with charter for a good part of last year well 3/4 of the year to be exact. after complaining so many times I got a new cable and phone modem still the same problem. I finely got Charter to send a supervisor technician and still same problem he gave me his number one friday morning then after oh 10 minutes of him looking over things his boss came out and gave me his card. told me to call him any time before 9pm i called him that very same day he takes and logged into to their stuff and call's me back oh 15 to 20 minutes later telling me that their was a total of 38 other modems both phone and internet have that very same problem. he gets the lines technicians out into our neighbor hood the following monday and so far the problem has been solved. so if you complain enough and make them check their lines outside not just the line coming to your house the main lines their very well could be a problem their in their lines. just saying is all

Collapse -
Depends on the server
by capoderra / December 13, 2013 7:57 PM PST

Netflix will have problems sending data simultaneously when everyone arrives home after work wanting to relax in front of the tv.

Collapse -
You Just Have Too Many Toys :)
by Hforman / December 14, 2013 2:37 AM PST

I do this thing for a living but it is going to be difficult to help in a forum like this (for me, at least). One thing: If you are running speedtest, I'd make absolutely sure that you delete your browser's cache before and after every test!.. (I've read that somewhere on at least one of these testing sites). If the targets are still in your cache, you may get false readings.

The best way to look at this is through special tools/meters such as from Fluke. For Internet access, I use my cable provider Time Warner and I pay the extra 50 cents per month for inside wiring. The best way I have found is to start at the Internet gateway and add things back in one at a time. But I'm used to using meters and special tools. If you can't resolve this, you may want to have a network specialist come out and take a look. Schedule him for the days and times when things are generally at their worst. Also, don't rule out that the websites are slow. They can be.

As someone correctly pointed out, your speed will be that of the slowest link between you and the website. That could be near the website or it could be satelite or something in your home. I remember that Internet access for some rural areas used to be by satelite service but it was slow becuase, what is really happening (pre-directv) is you send up a request for a web page and the response it just embedded in the flow of website pages coming down from the satelite and it was very slow. I don't know how the technology works today though.

Good luck.

Collapse -
Also...
by davidp007 / December 14, 2013 2:57 AM PST

The people responding here are pretty smart Wink I'm maybe not in their league, but will mention something.

TWC (TimeWarnerCable) in KY is moving away from speedtest.net. I agree. For me, it's not the best place to test my speeds, as I can get varying/inaccurate results. A tech said to use www.speakeasy.net/speedtest, and to use Chicago as the test city. Works great!!! I use that all the time now and no longer will use speedtest.net. It's off my list.

Of course, TWC usually wants people to test their speed at twc.com. I can understand that, and probably is the most accurate, seeing it is TWC's own server. But I mainly use speakeasy.net/speedtest for myself.

Cheers.

=======================

Collapse -
High speed bandwidth upgrade
by Craig Scranton / December 14, 2013 7:54 AM PST

Gosh, I pay 15.00 a month for the best bandwidth I can get from my ISP, and my down score is always in the high 80's, and my up is in the mid 30's. NOT cheap, but it is nice downloading and uploading. Craig Scranton, Mt. Arlington, NJ PS I do NOT have a commercial rate, just a home owner's rate.

Collapse -
WOW
by musicmugger / December 20, 2013 12:08 PM PST

You are a lucky boy, I pay $30 for 6meg where I live, and it's as reliable as a three legged race horse.

Collapse -
Bandwidth problems
by AdrianFox / December 20, 2013 4:32 PM PST
In reply to: WOW

Those complaining here should try having the service we get here in rural France. We have to pay 35 euros a month (no competition, only Orange can provide the service where we are!) and get a 512k connection, when we are lucky. From time to time the whole connection grinds to a total halt. I am guessing this is somebody else in the hamlet trying to download something? Most of the time I get a usable connection as long as I don't try anything that involves video.

I would complain even more bitterly were it not for the fact some of our neighbours are stuck with dial up!

The world is becoming divided into those with high speed and the rest of us in the Dark Ages!

Popular Forums
icon
Computer Help 47,885 discussions
icon
Computer Newbies 10,322 discussions
icon
iPhones, iPods, & iPads 3,188 discussions
icon
Security 30,333 discussions
icon
TVs & Home Theaters 20,177 discussions
icon
HDTV Picture Setting 1,932 discussions
icon
Phones 15,713 discussions
icon
Windows 7 6,210 discussions
icon
Networking & Wireless 14,510 discussions

Tech Tip

Tired of your tricky Wi-Fi password?

Stop trying to memorize a complicated sequence of numbers and letters. Learn how to change the default password.