dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
4368
share rss forum feed


The Limit
Premium
join:2007-09-25
Greensboro, NC
kudos:2
Reviews:
·Windstream
reply to telcodad

Re: [Speed] My Extreme 105 Speed Issues

I really am not trying to sound or be an asshat about this, but whose opinion are these speeds based on? Do you not see the problem with selling a service like this, and for the cost receive "anywhere from 0 to 105 megabit service."

Unless I see hard facts or policy stating at least a certain level and then up to a certain level, everything else is hearsay.
--
Do or do not, there is no try! - Yoda


KingKuei6

join:2002-11-18
Union City, CA
reply to Mike Wolf

Could be a router problem, assuming he's even using one, but then I suspect it probably is a gigabit connection anyway. To achieve 80Mbps download on Fast Ethernet 100Mbps would be very difficult even on a perfect day. Chances are he is connecting on a gigabit ethernet port, and something else in the router is the bottleneck.

Or maybe he's using a CAT5 cable, or a really bad CAT5e.



Mike Wolf

join:2009-05-24
Beachwood, NJ
kudos:3

Thats a very good analysis. Any ideas on what could be that bottleneck, or possible solutions?


bpratt

join:2006-10-24
Redwood City, CA

My first guess would be a performance limit in the device he's running the speed tests on. Could be the system bus, disk performance if caching is involved, CPU issues, etc, etc. especially since the OP is seeing the full upload speed.

My suggestion, which I think someone else made upstream, is run speedtest on multiple PCs on wired gigabit connections to your router at the same time, and see what their aggregate utilization up and down is. I'll bet that gets you close to 105/10, in which case you'll know that the original PC just can't pull down a full 105 Mbps on its own.

Bob



Mike Wolf

join:2009-05-24
Beachwood, NJ
kudos:3

ah that's a good idea. I'm curious, if the router itself could possibly be the bottleneck, and a replacement is an idea, which ones would be good?


bpratt

join:2006-10-24
Redwood City, CA

I seriously doubt the router is the issue. If it appears to be, after a multi-PC test, the best way to confirm is to use a local server and clients to test the router throughput, eliminating the Internet from the equation. But again, the router would be my last guess, unless it's a 10-15 year old box, which it almost certainly isn't.

If the OP does need a faster router, I don't have any specific recommendations, as I'm familiar with enterprise grade gear, and I think a Cisco 7600 would be overkill . But generally a couple hundred bucks more would buy plenty of performance, again in the remote case that the current router is the bottleneck.

Bob



Mike Wolf

join:2009-05-24
Beachwood, NJ
kudos:3

1 recommendation

true, but who really has a local server on hand? Anyway it would be a good idea to upgrade the router regardless, for various reasons such as improved user interface controls, IPv6 support, and overall increased hardware components. I mean the WRT54G which is kind of a staple in home networking over the years has some hardware revisions that max out at like 34Mbps on the WAN side even though it has a 10/100 port (from what I read on a networking site that does these tests) and think about this, a WRT54G V1 router with a Broadcom BCM4702 125Mhz processor, 16MB of ram, and 4MB of flash memory compared to an E4200 V2 router with a Marvell F6101AW 1.2Ghz processor, 128MB of ram, and 128MB of flash could perform similar, or wholly differently when introduced to the same extremely high speeds.


avlnch
Premium
join:2012-03-04
Boca Raton, FL
reply to avlnch

My plan is 105/10 and my best speed is 80/10. If this is normal, then I am happy with it. However, should I be getting higher speeds that are closer to my subscription? If so, I would really like to get there. If not, I am content with the speeds I am currently achieving. At 80/10 what Mbps download speeds should I be getting?

My Current Home Networking Setup:
Apple Airport Extreme (router)
Apple Airport Extreme (acting as bridge)
Apple Airport Extreme (acting as bridge)

I have my Airport Extreme Router and Arris WBM760 in a rack downstairs. I then have one Airport Extreme Bridge upstairs at the left side of the house, and have the other Airport Extreme Bridge upstairs at the right side of house. My two AEX Bridge's ARE hard wired down to the downstairs rack and directly into the AEX router.

I do have two Cisco/Linksys E4200 routers in the box. I was not positive on pulling the trigger on swapping out my Apple networking setup with my Cisco networking hardware.

If I did swap my router out with the Cisco E4200, should I just buy two bridges or should I buy a third E4200? Then I would have three and could put two in bridge mode. It would replicate my current setup however all hardware would be Cisco (not Apple).


bpratt

join:2006-10-24
Redwood City, CA

Before you worry about the routers, try the multiple PCs hard wired to the router hard wired to the cable modem and see where that gets you in terms of total bandwidth up and down. If you are getting 80/10 with one PC you are probably fine, especially if that's over wireless. That would be shockingly good wireless performance, and the wireless portion would be limiting you, not the cable portion.

Bob



Mike Wolf

join:2009-05-24
Beachwood, NJ
kudos:3
reply to avlnch

Well I know the E4200 has a bridge mode, and its my personal favorite for a router. I look at it as this, it can't hurt swapping everything over to the E4200's and try it out. If it works and everything is amazing then there you go, if it doesn't work you can always return them. Do you plan on going with the E4200 version 2 or the original version 1?


jbcc

join:2012-03-12
reply to avlnch

Good Afternoon avlnch,

I would open up a Comcast Direct post about your connection. I would ensure your system connects to an ATDMA(64QAM) US channel and not a TDMA (16QAM). You should be connected to the ATDMA channel. See if that improves your speed any.

Also, what kind of speed results do you receive if you plug directly into the modem without a router?

Thanks


avlnch
Premium
join:2012-03-04
Boca Raton, FL

1 edit
reply to avlnch

I have a poor speed update as of today (which I believe has been going on for a few days now). This speed test was taken at 12AM (midnight), and from a iMac computer wired directly to the router which is wired directly to my Arris D3 modem.

Wired Speed Test: 60/10 P=09

Latest ShaperProbe Test:

Office-iMac:~ avlnch$ /Volumes/ShaperProbe/ShaperProbe
DiffProbe Release. January 2012.
Shaper Detection Module.

Connected to server 217.163.1.87.

Estimating capacity:
Upstream: 10308 Kbps.
Downstream: 92129 Kbps.

The measurement will last for about 2.5 minutes. Please wait.

Checking for traffic shapers:

Upstream: No shaper detected.
Median received rate: 10080 Kbps.

Downstream: No shaper detected.
Median received rate: 87024 Kbps.



ARRIS MODEM INFORMATION: RF Parameters / Status / Interface Parameters / Event Log

DOWNSTREAM
DCID | Freq | Power | SNR | Modulation | Octets | Correcteds | Uncorrectables

Downstream 1 232 597.00 MHz -6.83 dBmV 36.17 dB 256QAM 16515758942 4284 6051
Downstream 2 229 579.00 MHz -6.73 dBmV 35.78 dB 256QAM 15382225139 1200 4434
Downstream 3 230 585.00 MHz -7.18 dBmV 36.17 dB 256QAM 16296319581 1113 4522
Downstream 4 231 591.00 MHz -7.46 dBmV 35.78 dB 256QAM 16032513617 1267 4623

UPSTREAM
UCID | Freq | Power | Channel Type | Symbol Rate | Modulation

Upstream 1 3 36.50 MHz 50.50 dBmV DOCSIS2.0 (ATDMA) 5120 kSym/s 16QAM

STATUS
System Uptime: 6 Days : 10 Hours : 00 Minutes
Computers Detected: staticCPE(0), dynamicCPE(1)
CM Status: OPERATIONAL
Time and Date: Sat 2012-03-17 23:00:00

INTERFACE PARAMETERS
Interface Name | Provisioned | State | Speed (Mbps) | MAC Address

LAN Enabled Up 1000(Full) 00:00:00:00:00:00
CABLE Enabled Up ----- 00:00:00:00:00:00

Modem Event Log
2/20 Entries Shown (Entries Since Last Reset)

Date: 03/13/2012
Time: 03:30
Event ID: 84000700
Event Level: [5]
Description: RCS Partial Service;CM-MAC=00:00:00:00:00:00;CMTS-MAC=00:00:00:00:00:00;CM-QOS=1.1;CM-VER=3.0;

Date: 03/15/2012
Time: 01:12
Event ID: 82000200
Event Level: [3]
Description: No Ranging Response received - T3 time-out;CM-MAC=00:00:00:00:00:00;CMTS-MAC=00:00:00:00:00:00;CM-QOS=1.1;CM-VER=3.0;


Okay, so I was reading over this post I am currently writing to just check it over before hitting submit. I saw that my shaper probe test conflicted with my speed test (from speedtest.net). I took MANY speed tests (about 7) and came to the highest result of 60/10. After the shaper probe results, I decided to go ahead and re-take the speedtest.net test.. and my speeds all the sudden jumped up (still on computer hardwired to router - to modem).

Speed Test Results Before Shaper Probe: 60/10 P=09
Speed Test 001
Speed Test 002
Speed Test 003
> 1GB/1024MB Test File Download (source): DL SPEED OF 950 KB/sec

==== SHAPER PROBE TEST TAKEN ====

Speed Test Results After Shaper Probe: 85/10 P=09
Speed Test 004
Speed Test 005
Speed Test 006
> 1GB/1024MB Test File Download (source): DL SPEED OF 2.3 MB/sec


So now I have a few questions, which I would like to ask all of you guys. Let me say this first.. I am very impressed with DSLR, especially the content available within the forums and the knowledge of DSLR members. I couldn't get better information anywhere else, and for that I truly appreciate and value your time and effort helping me with my connection issues. Thank YOU.

1. I am at a loss here as I do not understand how my subscription of 105/10 which I was able to obtain 80/10 out of it, has suddenly dropped down to 60/10!? It stayed at those low speeds for quite a few days, then when I do the shaper probe test, my connection speeds jump back to 80/10. What happened?

2. I see that Comcast offers two types of 105/10 packages.. You can get an XFINITY 105/10 plan or you can get a BUSINESS 105/10 plan. What is the difference? Are they two separate lines? So if I switched to a BUSINESS 105/10, would I be placed on a less active or better line and thus achieve better speeds?

3. What do you guys make of all this? I started writing to ask about why I dropped to 60/10, and now I am finishing this post asking about how I went back up to 80/10 after running one shaper probe test.


gar187er
I do this for a living

join:2006-06-24
Dover, DE
kudos:4
reply to avlnch

for all tests take your router out of the equation...which you failed to do...

1. noise, traffic, speedtest site overwhlemed....

2. different modems....your still connnected to the same plant......routing can also be varied however...this is not dsl....they dont swap pairs, or coax for that matter....same wires....

3. take the routers out the loop, and try again.....and also try real download tests....not speed test servers....
--
I'm better than you!


avlnch
Premium
join:2012-03-04
Boca Raton, FL
reply to avlnch

OK, I have new news as of this afternoon. I did repeat testing this afternoon and the results are interesting..

Just a recap of last night (the post before this). I completed a speed test and a file download. I achieved 60/10 and maximum download speed of about 900 KB/sec. Immediately after that, I completed a shaper probe test which came back saying 80/10 to my suprise, so I then did another speed test and file download test.. the results were back to normal at 80/10 and file download speed of 2.3 MB/sec.

Now this afternoon (the very next day) I completed a speed test and once again am at 50/10. I went ahead and did a Shaper Probe test which came back with DIFFERENT results than last night, here they are:

SHAPER PROBE TEST #1
Office-iMac:~ avlnch$ /Volumes/ShaperProbe/ShaperProbe ; exit;
DiffProbe release. January 2012.
Shaper Detection Module.

Connected to server 213.244.128.168.

Estimating capacity:
Upstream: 10017 Kbps.
Downstream: 58969 Kbps.

The measurement will last for about 2.5 minutes. Please wait.

Checking for traffic shapers:

Upstream: No shaper detected.
Median received rate: 9818 Kbps.

Downstream: Burst size: 32347-33067 KB; Shaping rate: 38453 Kbps.

logout
[Process completed]

SHAPER PROBE TEST #2
Office-iMac:~ avlnch$ /Volumes/ShaperProbe/ShaperProbe ; exit;
DiffProbe release. January 2012.
Shaper Detection Module.

Connected to server 4.71.254.149.

Estimating capacity:
Upstream: 10136 Kbps.
Downstream: 82356 Kbps.

The measurement will last for about 2.5 minutes. Please wait.

Checking for traffic shapers:

Upstream: No shaper detected.
Median received rate: 9781 Kbps.

Downstream: Burst size: 3719-4122 KB; Shaping rate: 66808 Kbps.

logout
[Process completed]


I just went ahead and completed another speed test after this afternnon's shaper probe and I am now at 70/10.

So my speed tests before shaper probe were 50/10. After shaper probe, my speed tests were coming back at 70/10.

I just went ahead and did an actual file download test, just like I did last night (from thinkbroadband) I only did this after the shaperprobe test, and the download speed was 2.2 MB/sec which is about identical to the download speed I was getting late last night when the shaper probe allowed my speeds to get back to 85/10 (temporarily as today/this afternoon it went back down).

This is the very first time I got a shaper detection. What exactly does this mean? Comcast HSI is throttling me? What can I do about this?


ExoticFish

join:2008-08-31
Stuarts Draft, VA

Shapers = Powerboost.



tshirt
Premium,MVM
join:2004-07-11
Snohomish, WA
kudos:4
Reviews:
·Comcast
reply to avlnch

said by avlnch:

This is the very first time I got a shaper detection. What exactly does this mean? Comcast HSI is throttling me? What can I do about this?

This part of the shaper test

Estimating capacity:
Upstream: 10136 Kbps.
Downstream: 82356 Kbps.
shows the line can still deliver 100+/80

The second part shows that Something---APPEARS to be limiting your results AT THAT MOMMENT.
Is it comcast? I doubt it.

But I would test again
clean boot, NOTHING else running on the computer (zillions of little updates, feed/email/ message/updaters running in the background can greatly skew the results) with a computer directly wired to the modem (even a decent laptop is fine, this test does not require much horsepower on your end). Run the test at least 6 times (you will get different servers) and when you run it Click the start button, remove your hand from the mouse and wait for results, do NOTHING for a minute (ok 30 seconds at least) and run again.

Obviously time of day, server load, CC network load, LAN, computer, nic and O/S are going to effect ANY sort of test, and sometimes don't mean much compared to real world results.


gar187er
I do this for a living

join:2006-06-24
Dover, DE
kudos:4

1 edit
reply to The Limit

said by The Limit:

I really am not trying to sound or be an asshat about this, but whose opinion are these speeds based on? Do you not see the problem with selling a service like this, and for the cost receive "anywhere from 0 to 105 megabit service."

Unless I see hard facts or policy stating at least a certain level and then up to a certain level, everything else is hearsay.

every service provider has the same wording, so that when some someone does a speedtest every day and has 99megs; he cant get credit for every day he doesnt see 100+.....its nothing new...
--
I'm better than you!