CoreXchange/Colounlimited Issues?

Portal Home > Knowledgebase > Industry Announcements > Web Hosting Main Forums > Providers and Network Outages and Updates > CoreXchange/Colounlimited Issues?

Posted by gordonrp, 06-24-2011, 02:44 AM
Anyone having issues at colounlimited/corexchange right now? One of our racks is having issues (major packet loss, but not loss between racks).
Posted by CN-Jeremy, 06-24-2011, 02:51 AM
Yes.. I am seeing packet loss and slow speed.. Even the ColoU and CoreXchange sites are suffering.
Posted by gordonrp, 06-24-2011, 02:54 AM
I've sent various documentation over the past hour without any positive feedback from them. Calling and asking to speak to the senior network guy results in "I don't have his direct number."

I don't see any issues on their websites, nor any of our servers in another rack.

Can you post a trace out from the servers you are having issues with?
Posted by CN-Jeremy, 06-24-2011, 02:58 AM
Looks like they are getting it resolved.. My sites look like they are returning to normal speed and the packet loss is almost gone completely.
Posted by gordonrp, 06-24-2011, 03:05 AM
Same here. It started at 1AM EST, we saw a 90% traffic drop then, and was just resolved around 2:57AM EST. This should of been fixed quicker, and they should of been aware of it before I opened a ticket.
Posted by CN-Jeremy, 06-24-2011, 03:10 AM
I didn't notice it till it got bad.. I had my other PC downloading a game from Steam for my son so my internet has been slow all night.
Posted by CN-Jeremy, 06-24-2011, 03:17 AM
I have the direct number for Michael Walti on the CoreXchange side. He is the one that complete most of the request i put in for network related stuff.
Posted by gordonrp, 06-24-2011, 05:01 AM
Apparently now fixed, total event 4 hours. Fingers crossed it's actually fixed now.
Posted by CN-Jeremy, 06-24-2011, 01:13 PM
This is the first time it has gone down in the 6 months I have been there so I guess that not to bad compared to other DC's that seem to have issues every few weeks.
Posted by gordonrp, 07-03-2011, 01:00 AM
You guys may be having issues again, I see the exact same issue happening again. Massive packetloss for a few hours for routes going via "vlan557" which last time they told me was one of their peers or uplinks. Going on for about 4 hours so far.

http://www.uploadscreenshot.com/image/376185/9421951
Posted by kris1351, 07-03-2011, 07:46 AM
Which DC were you seeing issues in. I never saw anything at 8600.
Posted by LoginTech, 07-03-2011, 01:11 PM

Quote:

Originally Posted by gordonrp

You guys may be having issues again, I see the exact same issue happening again. Massive packetloss for a few hours for routes going via "vlan557" which last time they told me was one of their peers or uplinks. Going on for about 4 hours so far.

http://www.uploadscreenshot.com/image/376185/9421951

That's not real loss, notice the lack of matching loss on subsequent hops? That's just the router shedding some ICMP requests.
Posted by FastServ, 07-05-2011, 09:48 AM

Quote:

Originally Posted by gordonrp

You guys may be having issues again, I see the exact same issue happening again. Massive packetloss for a few hours for routes going via "vlan557" which last time they told me was one of their peers or uplinks. Going on for about 4 hours so far.

http://www.uploadscreenshot.com/image/376185/9421951

There is no problem in that trace route. It basically shows zero packet loss
Posted by gordonrp, 07-05-2011, 11:32 AM
I appreciate that, however packetloss to final destination was an issue for a few hours. Also the latency generated on third hop was often over 100ms, same on the fourth (both in and out, screenshot is out only), affecting total performance. Network has become substandard in past few weeks, in my limited experience when the DC starts explaining issues with "low priority ICMP" it's a sign that things are overloaded in general. Our other rack that takes another route(r) is not affected.

This was the effect of the initial issue the other week (20kb/sec download speeds, 20%+ packetloss, datacenter ot aware of the issue until ticket raised), with packetloss on vlan557. This affected many clients, not just us: http://www.uploadscreenshot.com/image/379677/1058849

Still awaiting RFO for first outage. Network was solid first few months, with low latency and 0% packetloss along entire route, things are changing. Beware.
Posted by DMEHosting, 07-06-2011, 12:53 PM
Is this true, are you now on 'value' connections that will have issues from here on out? Can anyone confirm this with colounlimited?
Posted by gordonrp, 07-08-2011, 04:29 AM
It seems there was some miscommunication between them and us, which has now been corrected. Other than that first outage (which I understand wasn't handled optimally and they're keen to avoid), all should be acceptable for us now. Routing has changed, but it sounds like it is going to keep evolving (for the better) as they get more clients. Right now the routes are only 7%-10% slower (4-6ms) than our other rack and still hit Level3 on a lot of routes, so it's still very good bang for the buck. I don't see any lower grade carriers (e.g. cogent or HE) on any of the routes.
Posted by benj114, 07-16-2011, 11:17 PM

Quote:

Originally Posted by gordonrp

It seems there was some miscommunication between them and us, which has now been corrected. Other than that first outage (which I understand wasn't handled optimally and they're keen to avoid), all should be acceptable for us now. Routing has changed, but it sounds like it is going to keep evolving (for the better) as they get more clients. Right now the routes are only 7%-10% slower (4-6ms) than our other rack and still hit Level3 on a lot of routes, so it's still very good bang for the buck. I don't see any lower grade carriers (e.g. cogent or HE) on any of the routes.

I've not seen any major issues since moving to ColoUnlimited in Jan. Most of my routes either go Abovenet (98% of the time) or over IPTransit. The only issue I did have a few months ago was my ISP routing me to the London internet exchange and back..