dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
807
share rss forum feed


hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice

Terrible latency in the evenings... Powell, OH

This has been going on for the better part of a year with no real resolution. If there was an alternative to RoadRunner, I would be there by now.

I have had my modem swapped, all connections replaced, TV filter replaced. Nothing really resolves the problem. At one point, level 3 sent a request for the network team to split my node because they were seeing the problem with saturation. The request was put in, but action was never taken. That was August.

A tech came out on Wednesday, 11/14 and checked signals once again and claimed he would send an email to his supervisor about the node split. I know this is not a problem with my equipment or line to the pedestal because my neighbor 3 houses down sees the exact same problem. Latency in the 300-400ms range and packet loss.

This happens most nights, and is worst between 7pm and 10pm. Sometimes for a few minutes, other times up to an hour. I realize residential connections are susceptible to some slowdown, but .25Mbps down/up on a 30/5 package is just not acceptable. The last tech also said it takes about 30 calls in one area to justify a deeper look into the problem.

I have another tech scheduled for Tuesday from 8pm-9pm to hopefully have them present when the problem is exhibited, but I'm not holding my breath.

How do I get this escalated? I had worked with someone in the direct forum last spring and this summer, which got the node split request created. Also of note, they are building dozens of houses on my street alone, so load and saturation seems like a logical cause.



OSUGoose

join:2007-12-27
Columbus, OH

Try running Pingplotter to see what hop the slowdown is, if its the same as me, then yea I know its affecting us all. As I'm having similar issues on Insight/RR in NE Columbus, and getting the same run around.



hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice

1 edit

I always see it starting on the second hop (10.160.64.1), so I'm not even leaving the local RR network and getting latency. During the day, it's almost solid less than 50ms.

Tracing route to google-public-dns-a.google.com [8.8.8.8]
over a maximum of 30 hops:
 
  1     1 ms     1 ms    <1 ms  192.168.50.2
  2   592 ms    83 ms    69 ms  10.160.64.1
  3    98 ms    80 ms    21 ms  ae1-1112.cr02.clmcohib.midohio.rr.com [65.25.145.73]
  4   319 ms     *      353 ms  be21.clevohek-ccr01.mwrtn.rr.com [65.189.141.106]
  5   339 ms   485 ms   342 ms  ae10-0.cr0.dca20.tbone.rr.com [107.14.19.14]
  6    50 ms    45 ms   294 ms  107.14.19.135
  7   292 ms   268 ms   204 ms  66.109.9.66
  8   124 ms   123 ms    60 ms  209.85.252.46
  9    74 ms   261 ms   158 ms  72.14.236.148
 10    48 ms    60 ms   141 ms  72.14.238.70
 11   147 ms    50 ms    69 ms  216.239.49.149
 12    72 ms    36 ms    92 ms  google-public-dns-a.google.com [8.8.8.8]
 
Trace complete.
 


OSUGoose

join:2007-12-27
Columbus, OH

I'm getting the same, ranging 400ms while on RR.com, Notice I'm even hitting a local server, on the State of Ohio's network for universities:

Target Name: ohiodominican.edu
IP: 199.218.242.32
Date/Time: 11/16/2012 8:25:09 PM

1 2 ms [192.168.1.1]
2 1264 ms cpe-65-24-204-1.insight.res.rr.com [65.24.204.1]
3 627 ms gig2-18.instoh1-rtr01.insight.rr.com [65.24.192.77]
4 638 ms network-065-189-143-044.mwrtn.rr.com [65.189.143.44]
5 631 ms be21.clevohek-ccr01.mwrtn.rr.com [65.189.141.106]
6 659 ms ae10-0.cr0.dca20.tbone.rr.com [107.14.19.14]
7 635 ms ae-1-0.pr0.dca10.tbone.rr.com [66.109.6.165]
8 646 ms [107.14.16.82]
9 660 ms [64.57.21.234]
10 685 ms clmbn-r0-xe-4-0-1s100.core.oar.net [199.218.38.230]
11 653 ms clmbk-r9-xe-1-0-0s100.core.oar.net [199.218.38.26]
12 648 ms clmbs-r9-xe-0-0-0s100.core.oar.net [199.218.38.46]
13 * [-]
14 * [-]
15 * [-]
16 * [-]
17 * [-]
18 * [-]
19 * [-]
20 * [-]
21 * [-]
22 * [-]
23 * [-]
24 * [-]
25 * [-]
26 * [-]
27 * [-]
28 * [-]
29 * [-]
30 * [-]
31 * [-]
32 * [-]
33 * [-]
34 * [-]
35 * [-]

Destination not reached in 35 hops



hamburglar_

join:2002-04-29
united state

Our local routes look to be different, but I wonder if we are both seeing node congestion? Does yours slow down in the evening, or is it all day?



OSUGoose

join:2007-12-27
Columbus, OH

We take a slightly diff path, because im on the Insight/RR and your on TWC/RR, but we share:

5 631 ms be21.clevohek-ccr01.mwrtn.rr.com [65.189.141.106]
6 659 ms ae10-0.cr0.dca20.tbone.rr.com [107.14.19.14]

Beyond that we differ because I hit a local server, while you hit google's.

Try downloading Ping Plotter and let it run during a typical slowdown period, I've seen others get more results when RR is faced with that data.

And I get the normal evening slowdowns, but there are times off-peak I slow down too, and im on Turbo 15/.768 D2 package.



hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice

I see latency well before those hops. Generally

 2   592 ms    83 ms    69 ms  10.160.64.1
 3    98 ms    80 ms    21 ms  ae1-1112.cr02.clmcohib.midohio.rr.com [65.25.145.73]
 

If I ping 65.25.145.73 during slow times, I see the high latency before even getting past that router.


OSUGoose

join:2007-12-27
Columbus, OH

You need to download PingPlotter, it does pings in set intervals to normalize the results.



hamburglar_

join:2002-04-29
united state

I have it running against ohiodominican.edu now. I ran ping plotter off and on over the last 9 months, but it usually shows my first or second hop on the TW network, which would seem to be something more local to me...



OSUGoose

join:2007-12-27
Columbus, OH

ok, I just know from experience RR wont listen to the build in ping cmd, they wanted a Ping Plotter to see how it ranged. Could be your neighborhood needs rebalanced esp since its cold and cabling could be retracting from the fittings and allowing some leakage, or RR is being cheep and have oversold you.



hamburglar_

join:2002-04-29
united state

It happens in all temperatures, but was sporadic over the last couple months. It's been every night since Sunday so far this week. I'm leaning toward oversold, but good luck proving it.



bluepoint

join:2001-03-24
reply to hamburglar_

said by hamburglar_:

I always see it starting on the second hop (10.160.64.1), so I'm not even leaving the local RR network and getting latency. During the day, it's almost solid less than 50ms.

Tracing route to google-public-dns-a.google.com [8.8.8.8]
over a maximum of 30 hops:
 
  1     1 ms     1 ms    <1 ms  192.168.50.2
  2   592 ms    83 ms    69 ms  10.160.64.1
  3    98 ms    80 ms    21 ms  ae1-1112.cr02.clmcohib.midohio.rr.com [65.25.145.73]
  4   319 ms     *      353 ms  be21.clevohek-ccr01.mwrtn.rr.com [65.189.141.106]
  5   339 ms   485 ms   342 ms  ae10-0.cr0.dca20.tbone.rr.com [107.14.19.14]
  6    50 ms    45 ms   294 ms  107.14.19.135
  7   292 ms   268 ms   204 ms  66.109.9.66
  8   124 ms   123 ms    60 ms  209.85.252.46
  9    74 ms   261 ms   158 ms  72.14.236.148
 10    48 ms    60 ms   141 ms  72.14.238.70
 11   147 ms    50 ms    69 ms  216.239.49.149
 12    72 ms    36 ms    92 ms  google-public-dns-a.google.com [8.8.8.8]
 
Trace complete.
 

If you show that trace to a TW employee, they'll tell you there is nothing wrong with it since the destination is @92ms, ask hob. They will tell you to ignore the other hops and just look at the last hop. Of course, your problem is congested gateway(second hop) and you will need patience and time to convince TW that you have a congested node. Good luck!


hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice

I know what you are saying, and the problem with tracert is that it happens over multiple seconds while waiting for DNS lookups. When I can ping the third hop and my destination at the same time and I see very similar 300-400ms results simultaneously, it is hard to argue with.

The tech that came out this time did seem to accept that a bigger problem was at hand. This is especially so since I can confirm the same issue with my neighbor down the street. Hell, I was even over at her house running pings that showed the same results during one of the slowdowns.



bluepoint

join:2001-03-24

Keep up with the tech and don't keep your hope too high, sometimes they develop amnesia. Kidding aside, I hope they fix you up the soonest.



hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice
reply to hamburglar_

Click for full size
Pinging modem sub 2ms
Click for full size
Click for full size
Click for full size
Bandwidth usage during period
Last night and today have been crap. Tech was scheduled last wednesday, but when he called it wasn't acting up (of course). Here's what I'm seeing.


OSUGoose

join:2007-12-27
Columbus, OH

If you can try running ping plotter for a week, that way you may be able to see a trend when it acts up and schedule accordingly.


warmer20

join:2010-12-05
Dayton, OH
reply to hamburglar_

This has happened to me 6-7 times in the 3 years I've used TWC. You need to complain to tier 3 every day until you have a tech come to your house 3 (?) times. From there you can request to have the local special team assist you. They can actually fix your problem.



hobgoblin
Sortof Agoblin
Premium
join:2001-11-25
Orchard Park, NY
kudos:11
reply to bluepoint

"If you show that trace to a TW employee, they'll tell you there is nothing wrong with it since the destination is @92ms, ask hob."

Will you please stop running around these forums telling people what I might say.

Abnormal consistent ping times to the first hop indicates abnormal behavior.

Thanks.

Hob
--
"A foolish consistency is the hobgoblin of little minds."
- Ralph Waldo Emerson



hamburglar_

join:2002-04-29
united state
Reviews:
·ObiVoice
reply to warmer20

I was able to call tier 3 last night when it was really bad. He ran some ping tests from his end and said there was definitely a problem. They have been building a ton of houses in my neighborhood over the last year and the node is just overloaded. He also said they had multiple calls from my area last night... I guess 25% packet loss for hours finally pushed people to call in.