Ignight
02-08-2018, 04:45 PM
Hi Guys,
Here's what I found out today. This was over multiple phone calls and troubleshooting.
First: I started with Spectrum Cable. I spent over and hour on the phone with them detailing our common issue. This was that somewhere, we are losing signal and connection forcing us to time out or lag horrid.
While on the phone with them, I was reviewing the "Tracert" posted in the "Connection Issues" thread and the "Random Disconnect" thread and found a common source around the hop timeouts. I used both tracert images and sent them to technical support.
For me: "10ge2-9.core1.ash1.he.net" was a company in Fremont, CA called "Hurricane Electric". They are a 'Internet Backbone and Colocation Provider'. I called them up and to my surprise, the technician started running tests on his network looking for I/O issues. He found nothing wrong with his network but he identified the area where the issue is coming from. New York City.
He let me know they have approx 5 servers in NYC and that Project 1999 is on a Lightower network. His best guess was there is a transfer server between hurricane electric network and lightower network that is the root cause.
So I called Lightower. These guys did exactly what I expected. "I really can't help you without more data."
What data? Oh...a circut ID into Lightower or the new company name Crown Castle to identify where to look.
So, either Spectrum has the circut ID or Hurricane Electric has it, but Crown Castle/Lightower won't budge without this data.
This is as far as I got. Sorry I couldn't get any further today.
Here's what I found out today. This was over multiple phone calls and troubleshooting.
First: I started with Spectrum Cable. I spent over and hour on the phone with them detailing our common issue. This was that somewhere, we are losing signal and connection forcing us to time out or lag horrid.
While on the phone with them, I was reviewing the "Tracert" posted in the "Connection Issues" thread and the "Random Disconnect" thread and found a common source around the hop timeouts. I used both tracert images and sent them to technical support.
For me: "10ge2-9.core1.ash1.he.net" was a company in Fremont, CA called "Hurricane Electric". They are a 'Internet Backbone and Colocation Provider'. I called them up and to my surprise, the technician started running tests on his network looking for I/O issues. He found nothing wrong with his network but he identified the area where the issue is coming from. New York City.
He let me know they have approx 5 servers in NYC and that Project 1999 is on a Lightower network. His best guess was there is a transfer server between hurricane electric network and lightower network that is the root cause.
So I called Lightower. These guys did exactly what I expected. "I really can't help you without more data."
What data? Oh...a circut ID into Lightower or the new company name Crown Castle to identify where to look.
So, either Spectrum has the circut ID or Hurricane Electric has it, but Crown Castle/Lightower won't budge without this data.
This is as far as I got. Sorry I couldn't get any further today.