Hello,
We are reaching out because we noticed a significant increase in packet losses at the “nl-ams02a-rc2-lag-17-0.aorta.net” node (often 50%+) with our UPC connection at our office in Lausanne, when accessing the “us-east-1” region of Amazon Web Services.
To reproduce, run this from a machine located in the Lausanne area:
> mtr s3.us-east-1.amazonaws.com
The output will look as follows:
… SOME NODES
8. nl-ams02a-rc2-lag-17-0.aorta.net 68.9% 789 18.4 18.4 18.3 19.0 0.1
… SOME MORE NODES
The packet loss percentage is subject to some variations over the course of the day. For example, it was around 50% at 2:30am this morning, and then 0% at 3:30am, before coming back to the high percentage.
We happen to have some employees with a consumer UPC subscription in the Lausanne area, and they are experiencing the same packet losses as our office, also on the same “nl-ams02a-rc2-lag-17-0.aorta.net” node. This problem does not occur with employees who have other ISPs (including Swisscom, Sunrise, Salt).
Do you have some idea about what could be causing this issue and how to fix it?
Pierre