Closed MihaiSandor closed 1 year ago
marten-seemann
commented
1 year ago Have you taken a look at the dashboards? https://github.com/libp2p/go-libp2p/tree/master/dashboards
There's an AutoNAT and an AutoRelay dashboard that might give you some insight into what's going on.
MihaiSandor
commented
1 year ago Thank you for your response and suggestion.
I did take a look at the dashboards as suggested. Specifically, I examined both the AutoNAT and AutoRelay dashboards. I've also taken two screenshots for your reference:
Unfortunately, I was not able to gain much insight from these dashboards into why the AutoRelay is not detecting that the node is not publicly accessible and why it's not updating the addrs with the relay addr.
The metrics and visualizations provided in the dashboards were not very revealing or helpful to resolve the problem at hand.
marten-seemann
commented
1 year ago Looks like your node was only running for a few minutes, you need to give it some more time.
MihaiSandor
commented
1 year ago I understand your point about the runtime. However, even as we speak, my node has been running continuously for the past 30 minutes, and it's still facing the same issue.
Here are two new screenshots that illustrate the situation over this 30 minute interval:
Moreover, I would like to share that yesterday, I had the node running for more than 2 hours and the issue persisted. The AutoRelay still did not detect the node's inaccessibility and failed to update the addrs with the relay addr.
I am unsure of why this is happening and am hoping that there might be another reason for this issue that we haven't yet considered.
marten-seemann
commented
1 year ago Can you provide a screenshot of the swarm dashboard as well?
MihaiSandor
commented
1 year ago Certainly, I can provide a screenshot of the Swarm dashboard. Here it is:
Please note that in this screenshot, I've collapsed certain sections because they did not contain any data, and I wanted to focus on the sections that did contain data.
I'm still in the process of trying to fully understand this issue and appreciate your continued guidance and support. Let me know if there's anything else you'd like me to do or provide to help diagnose and rectify this problem.
marten-seemann
commented
1 year ago If the other sections don't contain any data, something is really going wrong. Apparently your node is not establishing any connections at all. Of course, AutoNAT and AutoRelay can't work then.
MihaiSandor
commented
1 year ago You're right, my node currently isn't establishing any connections. However, my main objective here is to validate and observe the functionality of AutoNAT and AutoRelay. I'm not trying to establish any additional connections to this node from other nodes at this moment.
From my understanding, even in the absence of other nodes attempting to connect, AutoNAT should be able to determine the node's accessibility. Since I'm on a network environment that is expected to make the node inaccessible, AutoNAT should fail and as a result, AutoRelay should automatically connect to the well-known relay that I've set up.
This has been the expected behaviour based on my understanding of the functionality. However, as seen from the dashboards and the node behaviour, this is not happening.
Could there be any other factors that I might be missing which are causing this issue? I'm eager to understand and fix this.
marten-seemann
commented
1 year ago AutoNAT uses existing connections to probe for your NAT status. If your node is just sitting idle and not establishing any connections, it won't have anything to work with.
MihaiSandor
commented
1 year ago That makes sense, and it provides a clearer picture of how AutoNAT operates. However, I've observed a different behaviour when I change that network to my home network. In that scenario, the list of addrs contains my public IP address and the respective port for each protocol. This occurs without making any extra connections to other nodes.
But when I switch to my 5G hotspot, this update does not occur. The node appears to behave differently in these two environments, despite no change in its connections.
Could this variation in behaviour between the ISP and hotspot environments point us towards any potential leads or explanations?
Description:
I am facing an issue where AutoRelay does not detect that the node is not publicly accessible (NAT'ed), and it does not update the addrs with relay addr even after 10-20 minutes.
Environment:
How to Reproduce:
Expected Behavior:
AutoRelay should detect that the node is not publicly accessible and update the addrs with the relay addr.
Actual Behavior:
Even after 10-20 minutes of execution, AutoRelay does not detect that the node is not publicly accessible and it never updates the addrs with the relay addr.
Program Log Addr Output:
Additional Context:
The static relay is set on a public instance. The relay used is from https://github.com/libp2p/go-libp2p-relay-daemon with version 0.4.0.
This is the first time I am facing such an issue, and I have followed the same steps as earlier without any issues.
Request:
I would greatly appreciate any guidance or insight on why this could be happening and how it can be rectified.
Version Information