Welcome to the Cox Community
Our community is a collaborative place to learn about Cox products and services. Jump into the conversation today!
Messages
Selected Messages
New Contributor
•
bad routing server causing high latency
Ok so i was wondering why i had always had the minimum of 50 ping to any server i tested never below 50 i didn't really think much f it till riot games creator of the game League Of Legends made a post of more centralized north american servers they left an ip to ping expecting to be in the 30~50 range of ping i pinged the server only to get results of 73 ping from my location i noticed people who live in Miami and other southern parts of Florida showing there tracerts of around 45-50 ping using isp's such as Comcast and centrylink keep in mind they are moving the servers from Portland Oregon to Chicago with me living in the panhandle of Florida going from 90 avg ping to 73 isnt a big change for how big the mileage from Portland to Chicago is anyways i was at a neighbors house who has centrylink i was testing his internet noticing i was getting around a average of 6 ping to a server approximately 120 miles away and i tested at my house only to get the standard 50 ms ping so after looking through tracerts i believe this server dalsbprj01-ae1.0.rd.dl.cox.net is the problem with the routing witch is located in Atalanta i was wondering if there is anyway you guys can fix/ or me blocking routing to this server and i pretty much am sure its this server because pinging servers more west such as in Dallas i can get under 50 ms ping when im not being routed threw this server. Btw i know cox is currently in a deal with riot games so you can use there pop's but this has nothing to do with other servers i have tested and got same results
edit also this server: ashbbprj02-ae1.0.rd.as.cox.net
ChrisL
Former Moderator
•
The thing to consider with routing is that any changes made affect all customers for better and worse. With that in mind our network engineers configure such aspects of the network with the best overall experience in mind. If you have any trace routes showing the problems you're experiencing we'd be happy to take a look and see if there is any feedback we could offer.
0
0
BrianS
New Contributor
•
I started doing some digging into this last night, and using the tool at http://www.monitis.com/traceroute/
I used the names given by johnkys, my own IP, and some other IPs, and I was running into this common intersection of multiple routes at Fort Walton Beach, FL. That left me confused, so I went to bed.
Now today, I'm finding it an extremely unusual coincidence that over in the "SB6141 unreliable connection" thread, there is this from a customer:
"Brand new Motorola SB6141 modem just purchased and installed last weekend and worked flawlessly until August 11.
Unreliable for about 3 days now. Started August 11. I had seen a Cox truck in my neighborhood in Ft Walton Beach and a few hours later noticed the lost connection."
0
0
johnkys
New Contributor
•
C:\Users\Mason>ping 104.160.131.3
Pinging 104.160.131.3 with 32 bytes of data:
Reply from 104.160.131.3: bytes=32 time=70ms TTL=53
Reply from 104.160.131.3: bytes=32 time=69ms TTL=53
Reply from 104.160.131.3: bytes=32 time=69ms TTL=53
Reply from 104.160.131.3: bytes=32 time=69ms TTL=53
Ping statistics for 104.160.131.3:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 69ms, Maximum = 70ms, Average = 69ms
C:\Users\Mason>tracert 104.160.131.3
Tracing route to 104.160.131.3 over a maximum of 30 hops
1 <1 ms <1 ms <1 ms 192.168.0.1
2 9 ms 8 ms 8 ms 10.18.0.1
3 13 ms 12 ms 11 ms ip68-1-11-214.at.at.cox.net [68.1.11.214]
4 11 ms 10 ms 12 ms ip68-1-11-214.at.at.cox.net [68.1.11.214]
5 54 ms 54 ms 54 ms ashbbprj02-ae1.0.rd.as.cox.net [68.1.0.254]
6 54 ms 54 ms 54 ms xe-0-1-3.er1.iad10.us.zip.zayo.com [64.125.12.165]
7 50 ms 48 ms 52 ms ae6.cr1.dca2.us.zip.zayo.com [64.125.20.117]
8 60 ms 61 ms 61 ms v11.ae29.cr1.lga5.us.zip.zayo.com [64.125.31.205]
9 59 ms 60 ms 60 ms ae10.mpr3.lga7.us.zip.zayo.com [64.125.20.77]
10 53 ms 53 ms 52 ms 128.177.169.158.IPYX-089814-970-ZYO.zip.zayo.com [128.177.169.158]
11 55 ms 53 ms 53 ms 104.160.132.35
12 69 ms 68 ms 68 ms 104.160.131.3
Trace completed
getting a constant 25-30 ping to alanta ga with centrylink cox = 50ms ping
i just want to know if theirs any way of not being routed through here or you to fix/uncongest this area the server im pinging is located in chicago
0
0
ChrisL
Former Moderator
•
That's probably the best route available. Those numbers are well within acceptable limits. With routing being based largely on geographical placement and peering relationships the path between two points will always differ from one provider to another.
0
0
BrianS
New Contributor
•
@Chris -
What about my "Fort Walton Beach" discovery? I find it real interesting that you have another customer in a different thread stating that they saw a truck in their area of Fort Walton Beach and shortly after that they started having problems.
Also, the "Pcola" thread is Pensacola, Florida, in the vicinity of Fort Walton Beach.
0
0
johnkys
New Contributor
•
getting routed through centrylink Atalanta point is 25-30ping
so basically you just telling me cox routing is bad
0
0
ChrisL
Former Moderator
•
Having looked at some of these I'm not finding enough similarities to suggest there is a larger issue. Should that change we'll definitely follow up further.
0
0
ChrisL
Former Moderator
•
Routing isn't something that can be simply be changed or described in terms of good or bad. No matter how a provider decides to route traffic the chosen routing will always be better for some types of communications and may not as idea for others. It's not much unlike our interstate highway system. Depending on where you want to go and where you are the ideal path may not be as ideal as if you were staring from a different point or going to a different destination.
0
0
johnkys
New Contributor
•
4. 11 ms 10 ms 12 ms ip68-1-11-214.at.at.cox.net [68.1.11.214]
5. 54 ms 54 ms 54 ms ashbbprj02-ae1.0.rd.as.cox.net [68.1.0.254]
0
0
ChrisL
Former Moderator
•
This would assume that us and Century Link use the same peering providers which is likely not the case. A lot of routing has to do with which peering providers are used by the Internet providers for the host on each end.
0
0
BrianS
New Contributor
•
Just for comparison purposes, this is what I get going to the same IP:
Tracing route to 104.160.131.3 over a maximum of 30 hops
1 1 ms 9 ms 9 ms 192.168.1.1
2 13 ms 18 ms 18 ms 10.64.80.1
3 12 ms 14 ms 14 ms 98.172.172.200
4 24 ms 28 ms 28 ms ashbbprj02-ae2.0.rd.as.cox.net [68.1.4.139]
5 27 ms 33 ms 38 ms xe-0-1-3.er1.iad10.us.zip.zayo.com [64.125.12.165]
6 33 ms 36 ms 29 ms ae6.cr1.dca2.us.zip.zayo.com [64.125.20.117]
7 41 ms 27 ms 31 ms v11.ae29.cr1.lga5.us.zip.zayo.com [64.125.31.205]
8 39 ms 37 ms 31 ms ae10.mpr3.lga7.us.zip.zayo.com [64.125.20.77]
9 36 ms 33 ms 31 ms 128.177.169.158.IPYX-089814-970-ZYO.zip.zayo.com [128.177.169.158]
10 38 ms 37 ms 40 ms 104.160.132.35
11 51 ms 58 ms 59 ms 104.160.131.3
Trace complete.
Ultimately, I end up with a 50-60ms ping myself, but it only reaches that point typically with the last couple of hops.
That said though, I am consistently hitting a different routing path at the peering point to zayo, meaning I am sent through a different set of equipment than john is on Cox's side. Once the traffic is handed off, we take the same route.
Interestingly enough though, I have found https://downdetector.com/status/cox-communications/fort-walton-beach
where someone claims that someone from within Cox said there is a software issue with a switch in Atlanta.
I wish you all would use IATA codes for the hops. For that first hop in zip.zayo, I know that IAD is Dulles International, so it is somewhere in the D.C. metro area. DCA is Reagan National. LGA is LaGuardia (NYC).
Edit: Or at least be immediately recognizable. I would think at.at is Atlanta, but I would believe that rd.as is Raleigh-Durham / Asheboro, NC. However, that Monitis site is saying Fort Walton Beach.
0
0
johnkys
New Contributor
•
I never said you were using the same peering providers I said the routing servers were almost identical geolocation wise.
0
0
johnkys
New Contributor
•
you live in fwb?
0
0
BrianS
New Contributor
•
No, I'm in Virginia. I just had done a little digging when you first posted and, while I possibly may have done this wrong, I suspected a problem in Fort Walton Beach, then the following morning, someone from there posted about seeing a truck and then their service went out.
Chris is right that the peering agreements could be different. Routes can also be different with Cox, because the system uses underlying metrics such as cost of routing through a particular path vs. another path.
I don't know where this "issue" is at, and if you will see any improvement with whatever Cox does, be that in Atlanta or elsewhere. 50ms is not bad at all, and it does not necessarily mean that actual traffic is slow, as routers can be told to respond to pings at a lower priority than other traffic (QoS).
For my situation, I am wondering why I have 9ms latency to my own router... That shouldn't be happening, as it is all inside this house.
0
0
johnkys
New Contributor
•
ya 9 ms to router is weird.
im fine with 50 ping to Atalanta but not 70 to Chicago when people in Cali are getting better results than me it gets frustrating.
if it were just to a website i wouldn't even be complaining but its to a game server making that 20-30 ms diff pretty big.
0
0