Client complaining about Connection Refused errors

skypanther

skypanther

Exalted Code Master!
I have a client complaining that they get "Connection was refused" errors frequently when trying to visit their web site. Yet they can connect fine sometimes. I never get this error. They can also connect fine to other sites (Google, other sites they own, etc.). They have tried Netscape, Internet Explorer, and Firefox and get the same issues in each.

Of course, they're a day's drive away and can't afford to have me come down to troubleshoot on site. They had a tech come in and disable the firewall in their wireless router. They claim to not be running firewall software on their PCs.

The site is hosted on wincf. I have checked the jodopulse graphs and don't see any outtages.

To me, this sounds like problems with their ISP selectively blocking sites. But, is there some other possibility that I should have them investigate?

I suppose I could open a ticket, but this seems more like user error or ISP problems than anything going on at Jodo.

Thanks,
Tim
 
connection refused? hrm I only see that on windows XP when I was running multiple apps and hitting the tcp max limits(it is sometihng like 10 connections, but i may be wrong)

Could they be running multiple apps, file sharing, etc at same time and hititng that?
 
Stephen,

Isn't that just an in-bound connection limit? Goodness, I have iTunes connected to radioio70s.com, Trillian connected to AIM & Yahoo, an IRC client, plus Firefox and IE with various open tabs all the time. I have a couple of file share connections open all the time. On occasion I have connections from my FTP client or HeidiSQL (mysql client) to a remote server. If anyone would hit limits on XP I think I'd be the one! :p

But, it's a good point and I will ask them.

Thanks,
Tim
 
I think the limit Stephen is talking about is the maximum number of TCP connections that can be opened at the same time, going through the handshaking process.
I've never encountered such a limit in Windows XP though, and I use P2P quite heavily.
Either way, as a total connection limit it would be ridiculously low. I currently have 625 open TCP connections :)
 
So, my client is able to connect to their site about 40% of the time. I got them to do a trace route. When it works, the trace route looks like this:

Code: 
Tracing route to www.renewableenergylongisland.org [204.14.107.243]
over a maximum of 30 hops:

  1     6 ms     6 ms     5 ms  10.33.96.1
  2     6 ms    12 ms    11 ms  dstswr1-ge3-1.rh.ehtnny.cv.net [67.59.224.161]
  3     8 ms    17 ms    11 ms  r1-ge0-2-2.mhe.hcvlny.cv.net [167.206.36.129]
  4     9 ms     7 ms     8 ms  r2-srp0-0.wan.hcvlny.cv.net [65.19.104.195]
  5     *        *        *     Request timed out.
  6     9 ms    10 ms    15 ms  rtr4-tg10-1.in.nycmny83.cv.net [64.15.0.17]
  7     *        *        *     Request timed out.
  8    11 ms    10 ms    10 ms  v3491.mpd03.jfk02.atlas.cogentco.com [154.54.7.9]
  9    10 ms    12 ms     8 ms  g10-0-0.core01.jfk02.atlas.cogentco.com [154.54.5.237]
 10    34 ms    34 ms    28 ms  p9-0.core01.atl01.atlas.cogentco.com [154.54.5.98]
 11    33 ms    38 ms    35 ms  p14-0.core01.jax01.atlas.cogentco.com [154.54.3.198]
 12    41 ms    41 ms    42 ms  p15-0.core01.mia01.atlas.cogentco.com [66.28.4.137]
 13    43 ms    43 ms    42 ms  p5-0.core02.mia03.atlas.cogentco.com [154.54.3.6]
 14    44 ms    42 ms    48 ms  vl3822.na21.b006657-0.mia03.atlas.cogentco.com [38.112.37.234]
 15    42 ms    45 ms    47 ms  Webhosting_dot_net.demarc.cogentco.com [38.112.7.230]
 16    46 ms    47 ms    47 ms  rtr1.m****here.biz [204.10.104.77]
 17    41 ms    41 ms    41 ms  204.14.107.243

Trace complete.

When it fails, it looks like the following.

Code: 
Tracing route to www.renewableenergylongisland.org [204.251.0.4]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  192.168.2.1
  2     5 ms     6 ms     6 ms  10.33.96.1
  3     6 ms     7 ms     5 ms  dstswr1-ge3-1.rh.ehtnny.cv.net [67.59.224.161]
  4    11 ms     7 ms     9 ms  r2-ge0-2-2.mhe.hcvlny.cv.net [167.206.36.133]
  5     9 ms    13 ms     9 ms  r1-srp9-0.wan.hcvlny.cv.net [65.19.104.209]
  6     *        *        *     Request timed out.
  7     *        *        *     Request timed out.
  8     *        *        *     Request timed out.
  9    12 ms    17 ms    18 ms  vlan89.csw3.NewYork1.Level3.net [4.68.16.190]
 10    14 ms    16 ms    17 ms  ae-83-83.ebr3.NewYork1.Level3.net [4.69.134.105]

 11    20 ms    18 ms    18 ms  ae-3.ebr3.Washington1.Level3.net [4.69.132.89]
 12    22 ms    17 ms    18 ms  ae-83-83.csw3.Washington1.Level3.net [4.69.134.170]
 13    16 ms    16 ms    16 ms  ae-81-81.ebr1.Washington1.Level3.net [4.69.134.137]
 14    30 ms    35 ms    39 ms  ae-2.ebr3.Atlanta2.Level3.net [4.69.132.85]
 15    44 ms    40 ms    33 ms  ae-78.ebr2.Atlanta2.Level3.net [4.69.134.53]
 16    30 ms    29 ms    29 ms  ae-21-54.car1.Atlanta1.Level3.net [4.68.103.98]

 17    30 ms    29 ms    29 ms  telcove-level3-10ge.Atlanta1.Level3.net [4.68.63.34]
 18    45 ms    44 ms    43 ms  g-1-1.c1-atln.telcove.net [24.56.97.18]
 19    42 ms    42 ms    44 ms  24.56.97.53
 20    42 ms    41 ms    42 ms  24.56.107.230
 21  64.66.101.90  reports: Destination host unreachable.

Trace complete.

They claim that it always fails at the same IP. When I look up 64.66.101.90 at dnsstuff.com, I see that it's managed by Telcove in PA.

Is this just a case of a bad DNS table at Telcove? If so, it should be a temporary problem. But this has been going on for a couple of weeks now.

I guess I don't know where to go from here in providing answers or possible resolution for my client.

Tim
 
Sounds like telcove routers are not working properly, but to be honest it is over/about 1000 miles away and I am not sure we can do anything to resolve it. it looks like when it tries to go internap it is failing, and only when it comes cogent does it work(which is a backup/standby connection that only works for people with an ISP forcing cogent, unless the internap is down).
 
Thanks for the reply Stephen. I didn't think you could resolve it. But this is beyond my area of expertise. I'm just hoping for some info that I can pass on to them.

I'll see if I can find a tech contact at telcove (or level3.com who seems to have bought telcove).

Tim
 
What about having them use OpenDNS? Could they possibly use that to better regulate their route through the internet?
 
Good thought. But, I'm thinking the best solution is to get telcove to fix their gear. My client's ISP serves a huge swath of Long Island so I'll bet the problem is affecting lots of folks.

Thanks,
Tim
 
I recommend he contacts his ISP. Usually they are helpful with these sort of routing issues. If the problem was closer to us, we could have tried working this out with InterNAP.
 
You must log in or register to reply here.
Back
Top