NTT high packet loss from US and BR to AU?

Anyone else notice this?

Or is this an AWS issue in APAC that hasn't been reported yet?

18. 72.0%

11. 71.4%

NJ/NYC to AU(aws)
9. 45.9% 772 10.1 16.4 9.2 94.4 13.3
10. 40.5% 772 69.6 72.7 69.3 149.2 9.0

Howdy all,

      I've been lurking for a long time, first time writing in. Please
excuse my inexperience. Javier, can you provide full traces, and
source/destination addresses?


Does it actually persist to your destination?

Loss in transit paths is simply ICMP de-prioritization unless it's losing packets all the way to the last hop.

Do you see any other indications of performance problems?


So we have a nagios box in the environment in Sydney and everything is 100%

new relic kept loosing connectivity to 30 plus servers on and off.

Guys from California can ssh in, some cant.

AWS reports everything operating as normal.

Guys from other parts of the world can and can't load web pages.

All servers show low usage (if you can ssh to them)

It seems to be getting better now but still not right.

This is from a box in AWS(sys) to level 3 dns server.

--- ping statistics ---
70 packets transmitted, 68 received, 2% packet loss, time 69744ms
rtt min/avg/max/mdev = 143.646/151.662/154.732/2.943 ms
[root@Webapp javier]#

Before it was 70% packet lost to that host. There is a mtr traceroute in a
previous email. Look for AU-US

Yup seeing the same. Following examples all show same loss pattern
between ~ 3:30 and ~ 4:30 UTC:

syd ntt - nyc ntt
syd ntt - mia ntt
syd ntt - cdg ntt (paris)
syd ntt - ams ntt

One example:


.-- My secret spy satellite informs me that at 2014-10-22 9:54 PM
Javier J wrote:

Thank you Andree,

This confirms it wasn't just us. I am curious if anyone knows what the
issue was. I can't find anything on NTT website.