Oregon Elastic IPs are Inaccessible; packet loss then completely inaccessible

Question

Routing to my Oregon-based AWS Elastic IP has gotten worse over the last few days. Now I can't reach the servers at all. I've tried associating a new elastic IP with my server and in worked for a few minutes. It's now failing again. Here's a trace route from my home connection at Spectrum.

The elastic IP I'm trying to reach is 54.184.62.53

```
traceroute to 54.184.62.53 (54.184.62.53), 30 hops max, 60 byte packets
 1  142-254-184-145.inf.spectrum.com (142.254.184.145)  10.004 ms  13.876 ms  14.704 ms
 2  lag-62.sndgcaxe01h.netops.charter.com (76.167.27.41)  27.170 ms  36.788 ms  37.047 ms
 3  lag-27.sndhcaax01r.netops.charter.com (72.129.2.86)  23.547 ms  28.553 ms  24.414 ms
 4  lag-22.lsancarc01r.netops.charter.com (72.129.1.0)  24.715 ms  24.738 ms  24.474 ms
 5  lag-26.lsancarc0yw-bcr00.netops.charter.com (66.109.3.230)  30.820 ms lag-16.lsancarc0yw-bcr00.netops.charter.com (66.109.6.102)  28.757 ms  29.637 ms
 6  lag-0.pr2.lax00.netops.charter.com (66.109.5.123)  26.283 ms  19.421 ms lag-800.pr2.lax00.netops.charter.com (66.109.7.225)  19.409 ms
 7  99.82.176.52 (99.82.176.52)  32.974 ms 99.83.70.220 (99.83.70.220)  15.882 ms  20.654 ms
```

...and that's where it dies. No matter which elastic IP is used, the trace always dies at 99.83.70.220.

I can ping 54.184.62.53 from my Verizon iphone connection. I can also reach the address by routing all of my Spectrum traffic through a ProtonVPN. But if the route passes through Spectrum/Charter, there's no connectivity.

Any help would be appreciated.

Answer

Hello Chad, thank you for your post. I am sorry to hear about the issues you observed when attempting to access the Elastic IPs attached to your server.

I do see that the packets sent by traceroute reach the Amazon network:

```
% whois 99.82.176.52 | grep -i -m2 ^org
organisation: ARIN
Organization:   Amazon.com, Inc. (AMAZO-4)
```

Despite the traceroute dying after reaching this host, this is not evidence of a routing issue. I was able to reproduce these results when using traceroute to evaluate the route from my home internet connection to an Elastic IP attached to an EC2 test instance. The last hop in my traceroute test was within the Amazon network, but there was no response from my Elastic IP. This was because traceroute uses UDP packets by default, and my EC2 instance's security group had no rules allowing inbound UDP connections. After I added a rule to my security group allowing UDP traffic, a subsequent traceroute test showed a successful connection to my Elastic IP.

With that said, I understand there may be another use case in which the connections you expect to work are failing. I would suggest you review the security group details for your EC2 instance to ensure the correct rules are in place.

If you would like assistance in reviewing the specific details of the issue, please feel welcome to submit a support case.

Answer

Hello chad_decker,

I am from the Premium Support Networking team. With regard to this issue, I would also encourage you to reach out directly via the support portal as this issue relates to specific resources under your account.
Raising a case via the support portal will allow us to dive a bit deeper into the issue and provide you with further information.

Oregon Elastic IPs are Inaccessible; packet loss then completely inaccessible

Relevant content