Interserver packet loss

Portal Home > Knowledgebase > Articles Database > Interserver packet loss

Posted by BELLonline, 09-09-2007, 01:04 PM
Is anyone else experiencing packet loss with interserver.net? It's pretty bad at the minute and was last night.
Posted by Dougy, 09-09-2007, 01:42 PM
Not noticing anything. Ping statistics for 66.45.228.100: Packets: Sent = 20, Received = 20, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 9ms, Maximum = 16ms, Average = 11ms
Posted by BELLonline, 09-09-2007, 01:51 PM
Hmm, it could be just for people in the UK or something - I have a server in Bluesquare UK and they are unable to maintain a 128kbps connection between them, and I also can't hold a 128k connection to my own connection without audio rebuffering. Others have noticed it as well. Their network is generally very good so I'm not really complaining about a day but was just interested to see if anyone else had the same problem. It's just a bit worrying because they aren't acknowledging that there is something wrong.
Posted by BELLonline, 09-09-2007, 01:57 PM
--- 66.45.246.xx ping statistics --- 24 packets transmitted, 22 received, 8% packet loss, time 23024ms rtt min/avg/max/mdev = 80.579/80.842/81.870/0.391 ms, pipe 2
Posted by astutiumRob, 09-10-2007, 06:22 AM
Nope, we have a server there from a recent takeover, and its constantly checked by both our UK and NL based monitorig systems and smokeping and there has been no appreciable (>1%) packet-loss since Interservers outage mid-August. Its going to be specific to the route(s) you are taking - try runnig an MTR and see what it shows.
Posted by BELLonline, 09-10-2007, 08:16 AM
It seems to have cleared up now (still not perfect) but yesterday and saturday it was only happening at peak time. It is strange. Most people are having no problems with buffering at all even in the UK but I had a really bad connection to the servers from my home connection as well as servers hosted in BlueSquare and RapidSwitch with around 10% packet loss. Using PingPlotter I could see that the problems started at algx and njiix, I hope it doesn't happen again tonight!
Posted by FastServ, 09-10-2007, 10:22 AM
Sounds like a peering problem, maybe between your ISP and one of their peers. Try sending Interserver a traceroute so they can reroute traffic (assuming they multi-home).
Posted by BELLonline, 09-10-2007, 10:59 AM
I've been in touch with them but they are not answering my tickets now. It seems strange that not everyone is experiencing the problem but I am having problems with the connection to my server at Bluesquare and also to my home internet connection - although they are completely unrelated, heres the traceroute for each: From Bluequare server: traceroute to 66.45.226.xx (66.45.226.xx), 30 hops max, 38 byte packets 1 ge-9-2.the.uk.euroconnex.net (217.112.88.1) 1.318 ms 1.188 ms 1.182 ms 2 gi4-40.telehouse-east2.core.enta.net (87.127.231.89) 1.237 ms 1.264 ms 1.192 ms 3 te4-4.telehouse-east.core.enta.net (87.127.236.97) 1.211 ms 1.215 ms 7.351 ms 4 core-01-lhr.njiix.net (195.66.225.20) 77.499 ms 77.451 ms 77.552 ms 5 core-02-teb1.us.njiix.net (64.20.32.218) 80.802 ms * 80.892 ms 6 core-03-teb2.us.njiix.net (64.20.32.234) 80.573 ms 80.799 ms 80.719 ms 7 edge-09-teb2.us.njiix.net (64.20.32.98) 81.132 ms 81.045 ms 81.025 ms 8 66.45.226.xx (66.45.226.xx) 80.757 ms 81.308 ms 80.643 ms From my home connection: Tracing route to 66.45.226.xx over a maximum of 30 hops 1 1 ms 1 ms 1 ms home [192.168.1.254] 2 52 ms 55 ms 54 ms l1.ar02.gs1.dsl.pipex.net [62.241.161.244] 3 50 ms 51 ms 50 ms ge-2-1-3.cr02.gs1.dsl.pipex.net [62.241.161.106] 4 54 ms 52 ms 54 ms pc9-104.cr05.tn5.bb.pipex.net [62.72.139.113] 5 52 ms 50 ms 50 ms 195.66.224.130 6 147 ms 149 ms 149 ms 65.106.3.73.ptr.us.xo.net [65.106.3.73] 7 146 ms 149 ms 149 ms p0-0-0.mar1.nyc-ny.us.xo.net [65.106.3.46] 8 139 ms 138 ms 137 ms ge4-0.CHR1.Secaucus-NJ.us.xo.net [71.5.189.130] 9 160 ms 141 ms 141 ms ip67-93-166-209.z166-93-67.customer.algx.net [67.93.166.209] 10 142 ms 141 ms 142 ms core-03-teb2.us.njiix.net [64.20.32.234] 11 143 ms * 142 ms edge-09-teb2.us.njiix.net [64.20.32.98] 12 141 ms 143 ms 141 ms 66.45.226.xx Trace complete. It looks to me like the njiix routers are overloaded at peak times because it has been fine during late night and early morning but starts at around 3PM uk time and goes on until about 2AM.
Posted by astutiumRob, 09-10-2007, 11:10 AM
both of those traceroutes go out from the isp's concerned, over linx, and through their peers network to the US. the enta(partial)-njiix will regularly be congested as one is massively oversold, and the other only has a small link to london the pipex-xo should be better, but you're on a residential connection, so being "pants" is normal
Posted by BELLonline, 09-10-2007, 11:28 AM
I'm surprised at that, I've been with Interserver for months now and haven't experienced these problems (maybe for an hour or two on a couple of occasions) I'm hoping that this isn't going to continue and I have to find another host, I've already had to leave one provider after their network went to **** recently I'm currently downloading a test file from my Interserver dedi, and am getting an average of 34KB/s on my server at Bluesquare (100Mbps) and 15KB/s to my home connection (8Mbps) - thats pretty crap! Last edited by BELLonline; 09-10-2007 at 11:33 AM.
Posted by dzone, 09-10-2007, 01:35 PM
We're using the NJIIX network and apparently there's a problem with a link to one of the NYC locations and they are working on it.
Posted by BELLonline, 09-10-2007, 01:42 PM
Yeah they just replied to my ticket and confirmed that there is an issue and they have gone to the location to take a look. "There is packet loss on our 1550 wavelength going to linx (others are unaffected)." It was quite a relief to see that they have found that there is a problem somewhere and hopefully it will be fully resolved before too long.
Posted by dzone, 09-11-2007, 09:56 AM
Looks like it's fixed on our end....
Posted by BELLonline, 09-11-2007, 11:01 AM
Yeah they have just confirmed that it has been fixed. Once they realised there was a problem it was fixed pretty quickly.
Posted by ZeroXcape, 09-13-2007, 06:42 PM
I'm still having huge problems. Both my server and their main site appear to be extremely slow right now. 50% packet loss. Pinging 66.45.228.100 with 32 bytes of data: Reply from 66.45.228.100: bytes=32 time=50ms TTL=48 Reply from 66.45.228.100: bytes=32 time=48ms TTL=48 Request timed out. Request timed out. Ping statistics for 66.45.228.100: Packets: Sent = 4, Received = 2, Lost = 2 (50% loss),
Posted by BELLonline, 09-13-2007, 07:19 PM
It's perfect from here: I've had no problems for a few days now. You might be best to open a support ticket with them.
Posted by astutiumRob, 09-13-2007, 11:57 PM
without seeing the route from you to that IP its imposisble to guess where the problem is. from here to Interserver over NAC to NJIIX to them is sub 1% packetloss 12 out of 10391 packets according to smokeping