
Tutor
•
14 Messages
AT&T Fiber Huge Latency and Packet Loss - Cumming, Georgia
Got AT&T Fiber some months ago and at first it was great. Within the last few weeks the service has seriously degraded to the point that it's nearly unusable. 2000+ millisecond pings and large packet loss most of the day and evening. Video constantly buffering. Zoom calls constantly dropping. I have attached a picture of the trace route from Ping Plotter and you can see the massive problems in the network. I know it is the external network because if I wake up in the middle of the night and test it, the latency and packet loss is non-existent. Can anyone at AT&T help?

JefferMC
ACE - Expert
•
33.5K Messages
1 year ago
If I am reading this correctly, you have nearly 1 SECOND latency to your in-home third-party Router (192.168.1.1). You cannot in any way say that problem is outside your home, or even related to your AT&T service.
I would not try looking at anything else until your in-home latency is 1 ms or less.
0
0
virago81
Tutor
•
14 Messages
1 year ago
Router is hard-wired to the AT&T device with a 1-foot CAT6. Again in the middle of the night there is no latency and packet loss.
Here's one where the latency is 74 milliseconds between the router and the AT&T device, but the downstream is still very bad. What do you make of that?
(edited)
0
0
JefferMC
ACE - Expert
•
33.5K Messages
1 year ago
100 "%PL" on a hop where hops beyond are responding is not a concern.
Note that using the term "packet loss" for a lack of an error response from a packet with a deliberately imposed error condition is a misuse of the term. The packet is intended to be "lost." The metric is measuring the percent of error responses returned. Note that it is not required that a hop provide the TTL-exhausted error, especially if it is busy doing other things (e.g. routing actual traffic).
Seeing as how hops beyond the 100 %PL hops (that you asked about) have about the same %PL as hop #2 (which is in your home), I again assert that a large part of your problem is before any traffic hits fiber.
0
0
virago81
Tutor
•
14 Messages
1 year ago
Here's one where the latency between the router and the AT&T device is 4.9 milliseconds. Downstream still very bad.
0
0
JefferMC
ACE - Expert
•
33.5K Messages
1 year ago
But your average latency to the Gateway (192.168.1.254) is still 150 ms. Something is going on between your router and the Gateway. Latency to your router is 4 ms. If Wi-Fi, then 4 ms is reasonable. The 150 ms ts not.
(edited)
0
0
virago81
Tutor
•
14 Messages
1 year ago
Ok. But what does that have to do with the huge packet loss and high pings downstream?
0
0
virago81
Tutor
•
14 Messages
1 year ago
Here everything is fine between the router and the device. Still high pings and PL downstream.
0
0
JefferMC
ACE - Expert
•
33.5K Messages
1 year ago
If a packet gets lost before it gets to your Gateway, it's not going to make it out of your house. So, if your first hop says 30% loss, you really shouldn't expect your 10th hop to do any better. The fact that your 2nd hop shows 36% and your 11th and 12th hops show an average of 35% tells me that you have an issue where about 35% of your packets aren't getting to the 2nd hop. Contrast this to your worry about the 100% packet loss at hops 3, 5, 9 and 10, where (by some incredible miracle if there were actually 100% packet loss) packets beyond them continue to respond at about a 65% rate.
Likewise, you're seeing a high ping rate inside your house. If it takes 0.910 seconds for a packet to get to your own router and back, and 1.8 seconds to get to your Gateway and back, by what miracle can you expect a packet to be any faster getting to the Internet???
0
0
JefferMC
ACE - Expert
•
33.5K Messages
1 year ago
What is it about your most recent plot (at about Noon today) concerns you? The 100 %PLs that are obviously not actually a problem because a hop responds beyond them. Is it the 12 ms latency at the destination?
0
0
virago81
Tutor
•
14 Messages
1 year ago
Ok. I made a few changes. Just to make sure, I changed the CAT6 cable from the router to the AT&T device. I noticed that the router had a hard-line out to a powerline adapter that wasn't being used, so I removed that.
Re-tested and the new graph looks much better. Thanks for the tips.
0