slow network access from internet

Customer Support Customer Only Discussion
1

Hi all,
currently I face slowness to all my customer domains on cluster1. Is there a incident at jodo or at pnap?

Ping responses are ok (for me around 130ms), but download-bandwith is very very limited. Only 8-20kbyte/sec instead of 300-800kbyte/sec.
Allready for more than 30 minutes now.

Thanks for anyone that has a clue.

Regards, Jan-Pieter

Tracerout from Europe/Netherlands/Amsterdam:
[SIZE=“2”][FONT=“Courier New”] 1 1 ms 1 ms 1 ms fritz.box [192.168.1.1]
2 18 ms 18 ms 18 ms lo0.ar16.ams3.network.solcon.net [212.45.35.205]
3 18 ms 19 ms 33 ms vlan502.crs1.ams1.network.solcon.net [212.45.45.1]
4 19 ms 18 ms 18 ms ge-2-3-0.104.ar1.AMS1.gblx.net [64.214.174.225]
5 115 ms 117 ms 117 ms INTERNAP.Tengigabitethernet2-2.ar1.MIA2.gblx.net [64.212.16.166]
6 117 ms 115 ms 116 ms border5.pc2.bbnet2.mia003.pnap.net [69.25.0.77]
7 117 ms 115 ms 116 ms webhosting-15.border5.mia003.pnap.net [216.52.162.66]
8 132 ms 133 ms 132 ms ibgp.border3.nota.mia.webhosting.net [64.71.226.34]
9 133 ms 132 ms 135 ms 64.71.225.26
10 133 ms 133 ms 135 ms 204.14.108.3[/FONT][/SIZE]

2

no issue that I am aware, I’ll put in a ticket to internap (pnap) and see if there is any impact.

The server you are on is not having any overall reduction in utilization of network port either.

It could possibly be route back to you, I am getting 175KB/s which is the max on the cable modem I am on at the moment.

3

Yea, it looks to be going back to you via cogent and that may be the issue…

Tracing route to lo0.ar16.ams3.network.solcon.net [212.45.35.205]
over a maximum of 30 hops:

1 4 ms 1 ms 1 ms 64.71.235.254
2 1 ms 1 ms <1 ms 64.71.225.25
3 1 ms <1 ms 1 ms gi1-18.ccr01.mia03.atlas.cogentco.com [38.112.7.
229]
4 1 ms 1 ms 1 ms te8-2.ccr02.mia01.atlas.cogentco.com [154.54.2.1
53]
5 27 ms 28 ms 27 ms te0-3-0-4.ccr22.atl01.atlas.cogentco.com [154.54
.3.25]
6 27 ms 27 ms 27 ms te0-4-0-7.ccr22.dca01.atlas.cogentco.com [154.54
.42.197]
7 28 ms 33 ms 28 ms te0-3-0-5.ccr22.iad02.atlas.cogentco.com [154.54
.41.238]
8 31 ms 28 ms 32 ms te2-1.ccr02.iad01.atlas.cogentco.com [154.54.31.
234]
9 28 ms 28 ms 28 ms kpn.iad01.atlas.cogentco.com [154.54.10.242]
10 59 ms 35 ms 34 ms nyk-s1-rou-1021.US.eurorings.net [134.222.228.9]

11 117 ms 117 ms 117 ms nyk-s1-rou-1001.US.eurorings.net [134.222.226.1]

12 117 ms 128 ms 117 ms asd2-rou-1022.NL.eurorings.net [134.222.226.169]

13 121 ms 121 ms * asd2-rou-1001.NL.eurorings.net [134.222.229.101]

14 112 ms 114 ms 116 ms 134.222.129.14
15 113 ms 113 ms 111 ms lo0.ar16.ams3.network.solcon.net [212.45.35.205]

Trace complete.

4

I am on 30 meg cable now, and an 8.8MB file download test downloaded in just seconds, so I am certainly not seeing it here, any update form your side?

I went through some things with internap and cogent but they said there did not seem to be any issue…

5

Hi Stephen,
After a few hours (approx 4-6 hours) the problem disappeared. By the way it was not related to the server I tracert’d, but also other servers/accounts had the same issue.
Downloadspeed is now back to 300 kbyte/sec which is reasonable. Still not the same speed as I get from more local servers, but this is not an issue.

What you say about that cogent could be the issue: is that a known issue? From your tracert it looks like jodohost is peering directly with cogent. Did you get info from Cogent what was causing the issue and how they are preventing this?
Reason I ask is that we see this bandwith-drop from week to week suddenly. Most of the time problem appears not long enough to report it, and also there is no evidence either that it’s jodo related because I can in those cases never isolate it to one jodoserver. Strange is also that when it happens the number of http-connections is not affected, so normally website load only slower, but they load. Only with large files the impact is very high.
Regards, Jan-Pieter