Is my ISP throttling my p2p bandwidth?

December 12, 2004 5:31 PM

p2pFilter: Is my ISP pulling a fast one on me? [mi]

My p2p lifeline (besides bittorrent) is eMule. eMule and I were in a blissful relationship. Flowers smelled better, the colors in life were more vibrant, and my transfers were going at a reasonable clip.

That was up until 3-4 days ago.

Having not touched anything on my computer or network setup, I noticed that the ping times for all the e2dk servers had gone up like crazy. Before they were 200 or so... Now, 800-5000. I'm not getting lowid's and I don't appear to be banned from anything on the servers' side. The rest of my net access seems about the same.

I know communication is the key to any good relationship. I'm wondering, was SBC DSL jealous of the love I had found with eMule? Did they try to sabotage my relationship? What can I do to get answers? Should I confront SBC DSL about it all? Has anyone else experienced this problem?

posted by drpynchon to Computers & Internet (14 answers total)

Run a traceroute to one of the e2dk servers.

On *nix: traceroute domain/ip_goes_here

I've not used Windows in many years, but I think: tracert domain/ip_goes_here

Anyways, you can look at the ping times it'll show to all the hops and see if it is one of your ISP's routers with high ping or something nearer to the e2dk server's end.
posted by brettcar at 5:42 PM on December 12, 2004

I've heard that ISPs are throttling especially heavy users, but don't know for sure. Did you overdo it?
posted by amberglow at 5:45 PM on December 12, 2004

Get into that router/modem and look at the port forwarding. Start here.
posted by cedar at 6:09 PM on December 12, 2004

Response by poster: I don't think this is a port forwarding issue. I've dealt with that substantially in the past with other routers and eMule. When the proper ports aren't forwarded, eMule connects with lowId, which I'm not getting. I've already forwarded the ports on my router previously to fix this problem and it was working great. The current problem arose without any changes to my ports.

Networking is not my specialty to be sure, and I've never run a tracert, so I have nothing to compare it to. Ran it on razorback2 and google as a control. Something is definitely different on the ed2k server but I'm not sure how to interpret the * and request timeouts. From my reading, the ping times through the ISP are comparable. Here's a capture of what I got:


Tracing route to razorback.ed2k.ch [195.245.244.243]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms 192.168.1.1
2 1372 ms 2332 ms 2115 ms adsl-xx-xxx-xxx-xxx.dsl.irvnca.pacbell.net [xx.xxx.xxx.xxx]
3 1074 ms 1148 ms 667 ms dist1-vlan50.irvnca.pbi.net [67.114.48.66]
4 879 ms 757 ms 947 ms bb1-g1-1-1.irvnca.sbcglobal.net [67.114.49.51]
5 975 ms 356 ms 579 ms core2-p5-0.cranca.sbcglobal.net [151.164.41.18]
6 207 ms 234 ms 161 ms core1-p8-0.cranca.sbcglobal.net [151.164.241.225]
7 762 ms 945 ms 646 ms core2-p3-0.crscca.sbcglobal.net [151.164.242.185]
8 897 ms 1608 ms 1834 ms bb1-p8-0.crscca.sbcglobal.net [151.164.40.62]
9 2259 ms 2278 ms 1978 ms ex1-p13-0.pxpaca.sbcglobal.net [151.164.190.206]
10 1464 ms 1475 ms 1763 ms zpr1.pao.cw.net [198.32.176.79]
11 1826 ms * 2163 ms so-6-0-0-ecr1.sfo.cw.net [195.2.3.77]
12 2052 ms 2485 ms 2072 ms so-7-0-0-ecr1.lax.cw.net [195.2.3.38]
13 2353 ms 2294 ms 2256 ms so-2-0-0-ecr1.was.cw.net [195.2.3.33]
14 2294 ms 1890 ms 2203 ms so-3-0-0-dcr1.was.cw.net [195.2.3.5]
15 1797 ms 1836 ms 1568 ms so-0-0-0-dcr1.par.cw.net [195.2.10.118]
16 1452 ms 1579 ms 1618 ms as0-dcr2.par.cw.net [195.2.10.162]
17 1937 ms 1168 ms 1892 ms so-3-0-0-bcr1.bru.cw.net [195.2.10.33]
18 1680 ms 1219 ms 1543 ms ae0-bcr2.bru.cw.net [206.24.147.30]
19 * * * Request timed out.
20 625 ms 1076 ms 802 ms ge0-0.iedgeixb2.isp.belgacom.be [194.78.0.61]
21 616 ms 536 ms 311 ms fe0-0.intl1.ixbelgium.legend.be [194.78.212.78]
22 1158 ms * 2484 ms razorback.ed2k.ch [195.245.244.243]

Trace complete.

-----------------------------------------------------------

Tracing route to www.google.akadns.net [216.239.63.104]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms 192.168.1.1
2 1733 ms 2187 ms 2544 ms adsl-xx-xxx-xxx-xxx.dsl.irvnca.pacbell.net [xx.xxx.xxx.xxx]
3 2422 ms 1393 ms 1909 ms dist1-vlan50.irvnca.pbi.net [67.114.48.66]
4 2504 ms 2634 ms 2421 ms bb1-g8-0-1.irvnca.pbi.net [67.114.49.17]
5 1796 ms 1523 ms 2023 ms core2-p5-0.cranca.sbcglobal.net [151.164.41.18]
6 1183 ms 1586 ms 1785 ms core1-p8-0.cranca.sbcglobal.net [151.164.241.22]
7 1546 ms 2056 ms 2080 ms core2-p3-0.crscca.sbcglobal.net [151.164.242.18]
8 2183 ms 1993 ms 1534 ms bb1-p8-0.crscca.sbcglobal.net [151.164.40.62]
9 2014 ms 2112 ms 1915 ms ex2-p14-0.eqsjca.sbcglobal.net [151.164.242.230
10 1719 ms 1746 ms 1978 ms ex1-p10-0.eqsjca.sbcglobal.net [151.164.191.66]
11 1666 ms 1376 ms 1315 ms eqixsj-google-gige.google.com [206.223.116.21]
12 1581 ms 1578 ms 1192 ms 216.239.48.210
13 1300 ms 1211 ms 1008 ms 216.239.49.12
14 2001 ms 1972 ms 982 ms 64.233.174.2
15 1617 ms 1871 ms 1373 ms 216.239.49.194
16 1504 ms 1364 ms 1543 ms 216.239.63.104

Trace complete.

-----------------------------------------------------------

Any other ideas?
posted by drpynchon at 6:17 PM on December 12, 2004

I know my friend has been having huge problems with eMule recently too (and it's probably not a problem on her side, as other p2p clients are working alright). Have you considered it might be some problem on eMule's side?

Sorry I can't be more helpful - neither of us is very computer literate.
posted by muddgirl at 6:40 PM on December 12, 2004

Woah, you are getting 1500ms pings to Google with a DSL service? Were you downloading/uploading when you did those traceroutes? Sounds like your DSL service sucks in general.

Comparision, last reachable hop to Google: 17 216.239.49.2 (216.239.49.2) 119.73 ms 116.484 ms 123.589 ms

Anyways, if the packet loss (those *'s) on razorback is the issue (probably, if Google is just as bad ping-wise) then it appears to be something on razorback's end, not your ISP. Either way, those pings are seriously funky. I had better ones back when I used dial-up. If you were downloading or uploading, try again, if you weren't then I'd switch ISPs.
posted by brettcar at 6:54 PM on December 12, 2004

Response by poster: Wait, sorry, bittorent was running... here's my pings without any transfers:

Tracing route to razorback.ed2k.ch [195.245.244.243]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 124 ms 127 ms 132 ms adsl-xxx.dsl.irvnca.pacbell.net [xxx]
3 149 ms 153 ms 156 ms dist1-vlan50.irvnca.pbi.net [67.114.48.66]
4 156 ms 150 ms 144 ms bb1-g1-1-1.irvnca.sbcglobal.net [67.114.49.51]
5 147 ms 146 ms 154 ms core2-p5-0.cranca.sbcglobal.net [151.164.41.18]
6 150 ms 146 ms 148 ms core1-p8-0.cranca.sbcglobal.net [151.164.241.225]
7 154 ms 155 ms 161 ms core2-p3-0.crscca.sbcglobal.net [151.164.242.185]
8 172 ms 175 ms 174 ms bb1-p8-0.crscca.sbcglobal.net [151.164.40.62]
9 150 ms 148 ms 145 ms ex1-p13-0.pxpaca.sbcglobal.net [151.164.190.206]
10 173 ms 176 ms * zpr1.pao.cw.net [198.32.176.79]
11 177 ms 174 ms 159 ms so-6-0-0-ecr1.sfo.cw.net [195.2.3.77]
12 184 ms 173 ms 173 ms so-7-0-0-ecr1.lax.cw.net [195.2.3.38]
13 212 ms 221 ms 224 ms so-2-0-0-ecr1.was.cw.net [195.2.3.33]
14 232 ms 230 ms 214 ms so-3-0-0-dcr1.was.cw.net [195.2.3.5]
15 312 ms 309 ms 309 ms so-0-0-0-dcr1.par.cw.net [195.2.10.118]
16 313 ms 311 ms 301 ms as0-dcr2.par.cw.net [195.2.10.162]
17 294 ms 308 ms 309 ms so-3-0-0-bcr1.bru.cw.net [195.2.10.33]
18 304 ms 312 ms 310 ms ae0-bcr2.bru.cw.net [206.24.147.30]
19 * * * Request timed out.
20 301 ms 305 ms 319 ms ge0-0.iedgeixb2.isp.belgacom.be [194.78.0.61]
21 310 ms 313 ms 323 ms fe0-0.intl1.ixbelgium.legend.be [194.78.212.78]
22 321 ms 321 ms 322 ms razorback.ed2k.ch [195.245.244.243]

Trace complete.

-------------------------------------------------

Tracing route to www.google.akadns.net [216.239.57.103]
over a maximum of 30 hops:

1 1 ms 1 ms 2 ms 192.168.1.1
2 161 ms 163 ms 158 ms adsl-xxx.dsl.irvnca.pacbell.net [xxx]
3 163 ms 158 ms 151 ms dist1-vlan50.irvnca.pbi.net [67.114.48.66]
4 144 ms 147 ms 155 ms bb1-g1-0-1.irvnca.pbi.net [67.114.49.1]
5 129 ms 135 ms 134 ms core2-p5-0.cranca.sbcglobal.net [151.164.41.18]
6 162 ms 160 ms 159 ms core1-p1-0.cranca.sbcglobal.net [151.164.241.221]
7 171 ms 173 ms 170 ms core2-p3-0.crscca.sbcglobal.net [151.164.242.185]
8 174 ms 177 ms 173 ms bb1-p8-0.crscca.sbcglobal.net [151.164.40.62]
9 156 ms 138 ms 126 ms ex2-p14-0.eqsjca.sbcglobal.net [151.164.242.230]
10 139 ms 150 ms 152 ms ex1-p10-0.eqsjca.sbcglobal.net [151.164.191.66]
11 163 ms 152 ms 154 ms eqixsj-google-gige.google.com [206.223.116.21]
12 177 ms 177 ms 172 ms 216.239.48.210
13 168 ms 172 ms 175 ms 216.239.49.168
14 171 ms 177 ms 175 ms 216.239.49.2
15 152 ms 144 ms 143 ms www.google.com [216.239.57.103]

Trace complete.
posted by drpynchon at 7:16 PM on December 12, 2004

Is BitTorrent -normally- running? It sounds like BT is saturating your upstream, which kills your internet connection and causes terrible pings.
posted by Jairus at 7:55 PM on December 12, 2004

What Jairus said. Otherwise, the packet loss is towards ed2k's end (hop 19 * * * Request timed out.)

I wouldn't blame your ISP.
posted by brettcar at 8:12 PM on December 12, 2004

Response by poster: Nah.. I rarely run both as I realize it taxes my system's resources (even moreso than my net connection). I get the ping problem (at least as listed through eMule) with BT off.. But yeah, I guess this is a downstream problem and out of my hands.
posted by drpynchon at 8:20 PM on December 12, 2004

If your BT is uploading too much, then your eMule might also be uploading too much. Try lowering your upload maximum in eMule by 50% and see if you have the same problem.
posted by Jairus at 8:23 PM on December 12, 2004

If you're getting 150ms on the first hop outside your house, something is wrong. That hop should be 10 or 20ms at most.

If you're certain that you don't have anything else running in the background and nobody sharing the connection, it may be your ISP. Try bypassing your router and connecting to your modem directly if you can, and also try rebooting the modem to see if that will help. If neither does, get the ISP on the line and have them investigate (which probably means sending out a tech). (And when you do call, don't say you're having problems with eMule -- say your webpages are loading slow, and let them investigate.)

Odds are good that this is just an overloaded link in your neighborhood, and they just need to upgrade capacity. If they really are throttling you, you'll find out when you call.
posted by Edge100x at 12:57 AM on December 13, 2004

Response by poster: Thanks guys, much appreciated.
posted by drpynchon at 7:55 AM on December 13, 2004

In your situation I'd be scanning my machine for viruses and trojans.

As a quick check, turn off all the networking apps you know you're using, and see if your cable modem stays busy.

I have yet to meet a hardcore p2p enthusiast with a properly secured machine.
posted by flabdablet at 4:20 PM on December 13, 2004