Forums » Bugs

Serious lag issues (once again)

12»
Sep 30, 2005 Gavan link
So yeah, pings of 400+ pretty constantly for the last few days. Traceroute attached (note, the first number, 2000+ is the result of little snitch).

Traceroute has started ...

traceroute to 207.67.118.34 (207.67.118.34), 30 hops max, 40 byte packets
1 d154-20-64-254.bchsia.telus.net (154.20.64.254) 2055.78 ms 23.373 ms 24.479 ms
2 208.181.240.62 (208.181.240.62) 24.364 ms 23.821 ms 22.621 ms
3 nwmrbc01gr01.bb.telus.com (154.11.4.66) 25.067 ms 24.804 ms 23.864 ms
4 nwmrbc01br01.bb.telus.com (154.11.10.53) 71.216 ms 71.48 ms 71.168 ms
5 edtnab02br01.bb.telus.com (204.225.243.2) 71.637 ms 71.604 ms 70.923 ms
6 wnpgmbabbr00.bb.telus.com (204.225.243.86) 72.11 ms 71.949 ms 72.679 ms
7 * * *
8 chcgildtgr00.bb.telus.com (154.11.11.18) 70.319 ms 71.252 ms 69.93 ms
9 peer-03-ge-2-0-0-34.chcg.twtelecom.net (66.192.252.101) 70.148 ms 70.708 ms 70.905 ms
10 core-01-so-0-1-0-0.chcg.twtelecom.net (66.192.244.21) 70.897 ms 71.557 ms 70.919 ms
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
Sep 30, 2005 Gavan link
hehe, oops... wrong ip address :P

the right one is forthcoming
Sep 30, 2005 Gavan link
There we go

Traceroute has started ...

traceroute to www.vendetta-online.com (204.11.209.34), 30 hops max, 40 byte packets
1 d154-20-64-254.bchsia.telus.net (154.20.64.254) 2257.06 ms 23.835 ms 23.613 ms
2 208.181.240.62 (208.181.240.62) 23.026 ms 22.875 ms 23.851 ms
3 nwmrbc01gr01.bb.telus.com (154.11.4.66) 24.833 ms 24.607 ms 24.197 ms
4 nwmrbc01br01.bb.telus.com (154.11.10.93) 76.138 ms 76.199 ms 76.076 ms
5 edtnab02br01.bb.telus.com (204.225.243.2) 75.459 ms 75.368 ms 75.096 ms
6 wnpgmbabbr00.bb.telus.com (204.225.243.86) 74.616 ms 74.9 ms 74.613 ms
7 toroonnlbr00.bb.telus.com (204.225.243.146) 75.096 ms 75.492 ms 74.847 ms
8 * * *
9 toroonxngr00.bb.telus.com (154.11.11.54) 74.311 ms 75.392 ms 73.859 ms
10 cogent.toroonxngr00.bb.telus.com (154.11.3.154) 74.101 ms 74.169 ms 74.091 ms
11 p15-2.core01.yyz01.atlas.cogentco.com (154.54.2.233) 76.816 ms 74.226 ms 73.86 ms
12 p13-0.core02.ord01.atlas.cogentco.com (66.28.4.213) 156.897 ms 168.079 ms 91.856 ms
13 p15-0.core01.ord01.atlas.cogentco.com (66.28.4.61) 88.401 ms 88.961 ms 89.058 ms
14 p5-0.core01.cle01.atlas.cogentco.com (66.28.4.221) 95.785 ms 94.884 ms 95.055 ms
15 p3-0.ca01.mke01.atlas.cogentco.com (154.54.1.46) 98.753 ms 347.216 ms 210.853 ms
16 38.112.26.190 (38.112.26.190) 97.583 ms 99.02 ms 96.998 ms
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
Sep 30, 2005 kihjin link
Was it just me or was the site down a few minutes ago?

Oh, and, I don't think www.vendetta-online.com and the game server share the same IP, so, if you are experiencing lag in game you'll need to traceroute that IP instead...
Sep 30, 2005 Beolach link
The production gameserver is majikthise.guildsoftware.com and the test server is zarniwoop.guildsoftware.com.

And if you're running Linux or some versions of Mac OS X, try using the -I option to traceroute, which tells it to use ICMP ECHO_REQUEST packets rather than UDP datagrams, and should get rid of all those * * * lines at the end.

$ /usr/sbin/traceroute -I majikthise.guildsoftware.com
traceroute to 204.11.209.42 (204.11.209.42), 30 hops max, 46 byte packets
<snip>
8 ae-0-0.bbr2.Denver1.Level3.net (64.159.1.114) 53.449 ms 52.846 ms so-4-1-0.bbr1.Denver1.Level3.net (209.247.11.70) 51.428 ms
9 ge-5-0-0.gar1.Denver1.Level3.net (4.68.124.218) 53.676 ms 52.108 ms 53.184 ms
10 ROADRUNNER.gar1.Denver1.Level3.net (4.71.40.2) 72.268 ms 71.516 ms 71.314 ms
11 pos4-0.milwwirtco-rtr1.wi.rr.com (24.94.160.18) 88.267 ms 89.127 ms 88.247 ms
12 srp14-0.milwwirtco-rtr2.wi.rr.com (24.160.225.2) 76.598 ms 76.636 ms 80.272 ms
13 srp2-0.milwwirtco-dar1.wi.rr.com (24.160.224.246) 89.488 ms 88.455 ms 89.880 ms
14 rrcs-67-52-10-170.west.biz.rr.com (67.52.10.170) 89.373 ms 90.983 ms 92.005 ms
15 204.11.209.42 (204.11.209.42) 90.870 ms 90.254 ms 89.398 ms
Oct 01, 2005 yodaofborg link
15 p3-0.ca01.mke01.atlas.cogentco.com (154.54.1.46) 98.753 ms 347.216 ms 210.853 ms <--- this is the problem, has been for months now :( My ping used to be a stable 110, and from the UK, thats pretty damn good, but since Anvil sold out, sometime it will be fine 110/130, sometimes I may as well not be here (ping of 400) Congent deny the fact that they have a flakey router, or that this is anything to do with them, but my traceroute shows...

2 8 ms 28 ms 7 ms oldh-t2cam1-a-ge-wan52-100.inet.ntl.com [80.5.16.13]
3 7 ms 6 ms 6 ms mant-t2core-a-ge-wan81.inet.ntl.com [213.104.24273]
4 6 ms 7 ms 7 ms man-bb-a-so-400-0.inet.ntl.com [213.105.172.241]

5 7 ms 7 ms 7 ms lee-bb-b-so-700-0.inet.ntl.com [62.253.185.194]

6 14 ms 8 ms 8 ms lee-bb-a-ae0-0.inet.ntl.com [62.253.187.185]
7 29 ms 14 ms 23 ms pop-bb-b-so-100-0.inet.ntl.com [62.253.185.238]

8 14 ms 41 ms 14 ms tele-ic-1-so-010-0.inet.ntl.com [62.253.185.82]

9 15 ms 16 ms 15 ms 212.250.14.62
10 84 ms 83 ms 83 ms p10-0.core01.bos01.atlas.cogentco.com [130.117.0
46]
11 99 ms 98 ms 103 ms p6-0.core01.cle01.atlas.cogentco.com [66.28.4.19
]
12 308 ms 108 ms 319 ms p3-0.ca01.mke01.atlas.cogentco.com [154.54.1.46]

13 109 ms 110 ms 125 ms 38.112.26.190
14 107 ms 107 ms 115 ms 204.11.209.42

(done in doze, cos i have been a naughty boy)
Oct 01, 2005 LeberMac link
OK, I'm in MILWAUKEE and I still have horrible pings on yoda's "problem" cogentco router: (154.54.1.46)

--- 154.54.1.46 ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max = 59.144/61.383/65.44 ms

--- 38.112.26.190 ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max = 32.524/35.776/47.747 ms

For reference my "direct" pings:
--- majikthise.guildsoftware.com ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max = 9.619/11.085/14.498 ms

--- zarniwoop.guildsoftware.com ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max = 26.887/29.456/33.836 ms
Oct 06, 2005 Beolach link
May or may not actually affect any of us, but I think it's worth a heads-up: Cogent and Level3, both Tier 1 (backbone) ISPs, have had "disagreements" and turned off their peering point. I used to go through both Level3 and Cogent on my way to majikthise.guildsoftware.com, but since Red Anvil (Guild Software's ISP) added the RoadRunner connections (sometime back around July or August, see the News page), I haven't gone through Cogent.

More info: http://ask.slashdot.org/article.pl?sid=05/10/05/2247207
Oct 06, 2005 LeberMac link
Those Bastards.
Oct 06, 2005 Screwball link
Oct 06, 2005 LeberMac link
Fracking BASTIDS!
Oct 06, 2005 tkjode link
Nov 01, 2006 Snax_28 link
So Cogent has once again begun to interfere with my gameplay. I'm not going to bother posting a traceroute because you need to examine a number of them to see the problem.

My question then is: Is there any way to manually re-route my connection to Guild? Or am I stuck with this crap?
Nov 01, 2006 chillum baba link
Just for fun:

traceroute to majikthise.guildsoftware.com (204.11.209.42), 64 hops max, 60 byte packets
1 dslrouter (192.168.1.1) 4.585 ms 0.598 ms 0.462 ms
2 72.235.60.1 (72.235.60.1) 28.151 ms 28.215 ms 27.787 ms
3 10.245.225.34 (10.245.225.34) 37.175 ms 28.965 ms 29.031 ms
4 sl-gw2-prl-6-2.sprintlink.net (144.223.63.85) 36.472 ms 28.704 ms 29.003 ms
5 sl-bb20-prl-0-0.sprintlink.net (144.232.30.21) 28.241 ms 28.547 ms 28.258 ms
6 sl-bb22-ana-10-1.sprintlink.net (144.232.8.206) 90.530 ms 90.768 ms 90.140 ms
7 sl-bb24-ana-14-0.sprintlink.net (144.232.1.161) 218.433 ms 277.794 ms 266.223 ms
8 sl-st21-la-11-0.sprintlink.net (144.232.8.92) 93.231 ms 92.992 ms 92.729 ms
9 p12-3.core01.lax05.atlas.cogentco.com (154.54.13.41) 101.305 ms 93.462 ms 93.059 ms
10 * * t3-1.mpd01.lax05.atlas.cogentco.com (154.54.6.186) 93.090 ms
11 t9-4.mpd01.lax01.atlas.cogentco.com (154.54.3.141) 93.423 ms 93.244 ms 93.347 ms
12 t8-4.mpd01.iah01.atlas.cogentco.com (154.54.2.197) 117.115 ms 117.695 ms 117.630 ms
13 t7-2.mpd01.dfw01.atlas.cogentco.com (154.54.1.73) 117.564 ms 117.449 ms 117.674 ms
14 t7-2.mpd01.mci01.atlas.cogentco.com (154.54.5.170) 126.512 ms 126.984 ms 126.455 ms
15 t2-1.mpd01.ord01.atlas.cogentco.com (154.54.5.174) 137.826 ms 137.539 ms 137.823 ms
16 v3489.mpd01.ord03.atlas.cogentco.com (154.54.5.18) 137.884 ms 138.340 ms 138.344 ms
17 g1-1.core01.mke01.atlas.cogentco.com (154.54.1.165) 140.260 ms 139.778 ms 140.377 ms
18 redanvil.demac.cogentco.com (38.112.26.190) 140.727 ms 139.851 ms 140.813 ms
19 majikthise.guildsoftware.com (204.11.209.42) 140.559 ms 140.597 ms 140.573 ms

EDIT: And a second run (wonder why hop 7 was so high on that one? And what's wrong with hop 10? besides that it's cogentco)

traceroute to majikthise.guildsoftware.com (204.11.209.42), 64 hops max, 60 byte packets
1 dslrouter (192.168.1.1) 1.341 ms 0.661 ms 0.516 ms
2 72.235.60.1 (72.235.60.1) 135.236 ms 70.780 ms 28.029 ms
3 10.245.225.34 (10.245.225.34) 30.623 ms 46.037 ms 28.833 ms
4 sl-gw2-prl-6-2.sprintlink.net (144.223.63.85) 28.439 ms 38.450 ms 28.087 ms
5 sl-bb20-prl-0-0.sprintlink.net (144.232.30.21) 31.139 ms 28.593 ms 28.305 ms
6 sl-bb22-ana-10-1.sprintlink.net (144.232.8.206) 90.649 ms 90.832 ms 91.525 ms
7 sl-bb24-ana-14-0.sprintlink.net (144.232.1.161) 91.054 ms 91.680 ms 91.590 ms
8 sl-st21-la-11-0.sprintlink.net (144.232.8.92) 92.730 ms 92.714 ms 93.047 ms
9 p12-3.core01.lax05.atlas.cogentco.com (154.54.13.41) 98.423 ms 92.984 ms 93.224 ms
10 * * *
11 t9-4.mpd01.lax01.atlas.cogentco.com (154.54.3.141) 93.151 ms 92.934 ms 92.631 ms
12 t8-4.mpd01.iah01.atlas.cogentco.com (154.54.2.197) 117.213 ms 117.136 ms 117.734 ms
13 t7-2.mpd01.dfw01.atlas.cogentco.com (154.54.1.73) 116.970 ms 117.858 ms 117.231 ms
14 t7-2.mpd01.mci01.atlas.cogentco.com (154.54.5.170) 127.080 ms 126.780 ms 126.515 ms
15 t2-1.mpd01.ord01.atlas.cogentco.com (154.54.5.174) 137.940 ms 137.657 ms 137.660 ms
16 v3489.mpd01.ord03.atlas.cogentco.com (154.54.5.18) 138.494 ms 137.822 ms 140.327 ms
17 g1-1.core01.mke01.atlas.cogentco.com (154.54.1.165) 140.186 ms 143.531 ms 140.243 ms
18 redanvil.demac.cogentco.com (38.112.26.190) 140.206 ms 140.308 ms 140.067 ms
19 majikthise.guildsoftware.com (204.11.209.42) 151.455 ms 140.283 ms 140.608 ms
Nov 01, 2006 Snax_28 link
That's exactly what mine looks like. Let's say hops 6-10 are cogentco. First trace, hop 7 will be upwards of 200 when everything else is fine. Second trace, no problems. Third trace hop 8 or 6 will be >200. Fourth trace, hop 9 will be the high one.

Anyway, basically makes gameplay unpredictable. I'll be dodging, dodging... all of a sudden I'm a sitting duck.

Sucks.
Feb 19, 2007 Aramarth link
I used to be a decent pilot. Then on January 30th my ping quadrupled. I haven't defeated anyone decent since... then a friend showed me this thread. My two cents:

traceroute to majikthise.guildsoftware.com (204.11.209.42), 64 hops max, 40 byte packets
1 n130s001 (204.111.130.1) 18.564 ms 4.911 ms 1.188 ms
2 10.100.100.33 (10.100.100.33) 1.623 ms 1.534 ms 4.045 ms
3 n169s097 (204.111.169.97) 1.952 ms 1.749 ms 1.884 ms
4 n169s065 (204.111.169.65) 4.660 ms 2.809 ms 2.627 ms
5 n169s233 (204.111.169.233) 9.642 ms 2.997 ms 6.806 ms
6 204.111.27.33 (204.111.27.33) 9.638 ms 11.346 ms 10.151 ms
7 * p4-0-1-0.r02.asbnva01.us.bb.gin.ntt.net (129.250.11.121) 10.715 ms 10.605 ms
8 p16-2-1-2.r21.asbnva01.us.bb.gin.ntt.net (129.250.2.62) 18.089 ms 10.988 ms 13.968 ms
9 t8-4.mpd01.iad01.atlas.cogentco.com (154.54.10.161) 28.713 ms 10.671 ms 11.769 ms
10 v3495.mpd01.dca02.atlas.cogentco.com (154.54.5.61) 10.377 ms 16.278 ms 11.421 ms
11 v3491.mpd01.dca01.atlas.cogentco.com (154.54.2.181) 13.898 ms 12.041 ms 12.222 ms
12 t2-4.mpd03.jfk02.atlas.cogentco.com (154.54.6.14) 47.057 ms 48.037 ms 111.308 ms
13 t7-2.mpd01.bos01.atlas.cogentco.com (154.54.5.242) 126.316 ms 54.591 ms 45.987 ms
14 t8-4.mpd01.ord01.atlas.cogentco.com (154.54.7.81) 45.298 ms 45.363 ms 44.354 ms
15 vl3490.mpd01.ord03.atlas.cogentco.com (154.54.6.206) 44.757 ms 44.459 ms 44.277 ms
16 g1-1.core01.mke01.atlas.cogentco.com (154.54.1.165) 46.646 ms 46.861 ms 46.472 ms
17 redanvil.demac.cogentco.com (38.112.26.190) 46.151 ms 60.113 ms 46.629 ms

A second attempt was identical, except for an asterisk at hop 10:
10 v3495.mpd01.dca02.atlas.cogentco.com (154.54.5.61) 10.320 ms 17.521 ms *

I'm assuming that means a dropped packet, because it hung there for a while. Seriously, is there any way to fix this? Losing 40% of my armor and never seeing the shots seriously makes the game less enjoyable.

There was one fight in particular.. I ended up a thousand meters from the guy I was fighting. He and I weren't the only ones there either, a third guy who was watching typed a WTF into sector chat as we did.
Feb 19, 2007 greengeek link
There's pretty much nothing you can do to change the route your packets take through the internet other than: switch ISPs (which might not help, or make it worse), move to another state, or offer sacrifices to the dark gods.

None are guaranteed to help, especially when the issue is almost always Cogent, who happen to be one of the 2 backbone providers for VO's Co-Lo host. Something tells me Cogent are the types who frown upon concepts like "network neutrality". Odds are they are trying some traffic shaping within their network, or they have one particular router along nearly everyone's path to majikthise which is routinely lighting on fire.
Feb 21, 2007 chillum baba link
erm... You are complaining about a ping of 46-47? (Though the last hop is missing, it's looking like it wouldn't be more than another ms.) If you can't "defeat anyone decent" with that ping... well... it's not lag that's the issue maybe.

Here's my current traceroute for comparison (and I defeat "decent" players all the time):

traceroute to majikthise.guildsoftware.com (204.11.209.42), 64 hops max, 60 byte packets
1 dslrouter (192.168.1.1) 0.777 ms 0.537 ms 0.471 ms
2 72.235.60.1 (72.235.60.1) 27.603 ms 28.116 ms 27.800 ms
3 10.245.225.34 (10.245.225.34) 28.564 ms 29.351 ms 28.519 ms
4 sl-gw2-prl-6-2.sprintlink.net (144.223.63.85) 28.025 ms 27.973 ms 28.002 ms
5 sl-bb20-prl-0-0.sprintlink.net (144.232.30.21) 28.048 ms 28.158 ms 28.017 ms
6 sl-bb22-ana-10-0.sprintlink.net (144.232.8.158) 82.772 ms 82.987 ms 82.979 ms
7 sl-bb24-ana-14-0.sprintlink.net (144.232.1.161) 96.068 ms 217.707 ms 89.902 ms
8 sl-st21-la-11-0.sprintlink.net (144.232.8.92) 91.825 ms 92.201 ms 92.094 ms
9 p12-3.core01.lax05.atlas.cogentco.com (154.54.13.41) 92.106 ms 92.467 ms 92.353 ms
10 t3-3.mpd01.lax05.atlas.cogentco.com (154.54.3.21) 92.072 ms 92.296 ms 91.830 ms
11 t8-2.mpd01.lax01.atlas.cogentco.com (154.54.6.225) 92.602 ms 92.412 ms 93.075 ms
12 t8-4.mpd01.iah01.atlas.cogentco.com (154.54.2.197) 135.995 ms 136.338 ms 136.243 ms
13 t7-2.mpd01.dfw01.atlas.cogentco.com (154.54.1.73) 136.448 ms 136.154 ms 136.454 ms
14 t8-1.mpd01.mci01.atlas.cogentco.com (154.54.5.14) 137.887 ms 343.619 ms 216.765 ms
15 t7-1.mpd01.ord01.atlas.cogentco.com (154.54.6.157) 136.207 ms 136.012 ms 135.683 ms
16 vl3490.mpd01.ord03.atlas.cogentco.com (154.54.6.206) 136.000 ms 136.968 ms 135.962 ms
17 g1-1.core01.mke01.atlas.cogentco.com (154.54.1.165) 137.903 ms 137.462 ms 137.718 ms
18 redanvil.demac.cogentco.com (38.112.26.190) 138.370 ms 138.634 ms 138.410 ms
19 majikthise.guildsoftware.com (204.11.209.42) 137.872 ms 137.912 ms 137.667 ms

And a second run:

traceroute to majikthise.guildsoftware.com (204.11.209.42), 64 hops max, 60 byte packets
1 dslrouter (192.168.1.1) 1.189 ms 0.531 ms 0.470 ms
2 72.235.60.1 (72.235.60.1) 28.166 ms 28.059 ms 39.129 ms
3 10.245.225.34 (10.245.225.34) 28.778 ms 28.773 ms 28.874 ms
4 sl-gw2-prl-6-2.sprintlink.net (144.223.63.85) 28.042 ms 28.064 ms 27.742 ms
5 sl-bb20-prl-0-0.sprintlink.net (144.232.30.21) 28.626 ms 27.808 ms 27.783 ms
6 sl-bb22-ana-10-0.sprintlink.net (144.232.8.158) 82.999 ms 82.845 ms 82.499 ms
7 sl-bb24-ana-14-0.sprintlink.net (144.232.1.161) 89.933 ms 89.943 ms 89.598 ms
8 sl-st21-la-11-0.sprintlink.net (144.232.8.92) 92.128 ms 91.886 ms 91.948 ms
9 p12-3.core01.lax05.atlas.cogentco.com (154.54.13.41) 92.351 ms 92.389 ms 92.094 ms
10 t3-3.mpd01.lax05.atlas.cogentco.com (154.54.3.21) 92.099 ms 92.188 ms 92.166 ms
11 t8-2.mpd01.lax01.atlas.cogentco.com (154.54.6.225) 92.317 ms 92.332 ms 92.349 ms
12 t8-4.mpd01.iah01.atlas.cogentco.com (154.54.2.197) 136.275 ms 136.285 ms 135.926 ms
13 t7-2.mpd01.dfw01.atlas.cogentco.com (154.54.1.73) 136.225 ms 136.243 ms 135.930 ms
14 t8-1.mpd01.mci01.atlas.cogentco.com (154.54.5.14) 136.202 ms 135.952 ms 135.997 ms
15 t7-1.mpd01.ord01.atlas.cogentco.com (154.54.6.157) 135.917 ms 135.565 ms 136.155 ms
16 vl3490.mpd01.ord03.atlas.cogentco.com (154.54.6.206) 136.103 ms 135.539 ms 135.414 ms
17 g1-1.core01.mke01.atlas.cogentco.com (154.54.1.165) 137.405 ms 137.629 ms 138.667 ms
18 redanvil.demac.cogentco.com (38.112.26.190) 138.491 ms 138.450 ms 138.865 ms
19 majikthise.guildsoftware.com (204.11.209.42) 138.005 ms 137.632 ms 137.444 ms
Feb 21, 2007 bojansplash link
Feb 22, 2007 incarnate link
Since this seems to be a common misconception, a couple of points about interpreting traceroutes.

A traceroute is a way of determining the path or route that your packets take to a given host (not the route the host takes to you, which is another problem).

The latency (pingtime) for any given point on the trace is only useful if all other pingtimes after that point are similarly high. For instance, if you saw 90ms pings for the first five hops, and then 1500ms pings for the five after it.. it's likely that there's a problem on the link between hop 5 and 6, or with the equipment at one of those points.

However, if you see high "pings" followed by low pings.. that means.. more or less nothing. This is because routers are designed to route packets, not respond to directed packets. They focus all their attention on routing packets that go *through* them, and often little to no time at all in responding to your pings.. which are really not relevant to the efficiency of the router.

Thus, only the point-to-point ping time (from you to majikthise, for instance) is useful in measuring actual latency. Pinging some
interim transit router is largely irrelevant.

Unfortunately, the most common problem to plague our users is one of asymmetric routing (two different paths being taken.. A->B and B-> don't follow the same route.. or even the same provider). This can happen for a number of reasons, a big one being individual providers padding out their route preferences to spread their bandwidth usage around. Unfortunately, it can lead to situations where you can get to us just great, but we can't get to you very quickly, or visa versa. A traceroute doesn't show this, and a ping will only show that it's "slow", it requires tracing from both ends to be useful.

And, even if you do diagnose the problem, it's non-trivial to solve. Those route pads are usually in there because if they weren't, a given peering point or transit link would become so saturated that it would no longer be useful.

Anyway, all things considered, it's amazing the Internet works at all most of the time. Our hosting, for the most part, is pretty good, and our connections are pretty fast.

As a final point, the number of "hops" in a path is really kind of irrelevant as well. Only the throughput and latency count. Different network technologies yield more or less hops.. these days I think people are mostly using Packet Over Sonet or GigE, which results in more visible interfaces (the "hops"). Older technology, like Asynchronous Transfer Mode, yielded less visible "hops", but wasn't necessarily better (the merits of a cloudy vs straight point-to-point network mesh being beyond this discussion).