-
Notifications
You must be signed in to change notification settings - Fork 452
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DHTDiscoveryCommunity IPv6 support #8276
Comments
The DHT community has its own separate pool of peers. It is expected that the peers in other communities do not appear in the DHT community. However, it is not expected that this community never gets any peers. Currently, the DHT community is not actually used by any Tribler components, making this a lower priority bug. I'll schedule this for 8.1.0. |
@egbertbouman I assigned you, because as the author of the DHT community you can authoritatively determine if there is indeed something wrong. |
I am seeing the following tracebacks for IPv6:
|
Hello I see a lot of equivalent errors in the logs of the docker container "image: ghcr.io/tribler/tribler:latest" installed a week ago ...
it never seems to involve private IP addresses. I have a feeling it has something to do with this issue, but maybe not. Thx for your help |
This issue seems to be about multiple things: too many mapped IPv4 addresses, the DHT not working, and peer discovery tracebacks in the The following seems to be happening. We receive an old-style introduction-request from a peer we haven't met before. The introduction-request has To me, this looks more like a general IPv8 issue. I've been looking at the tracebacks for a while now, and I don't really understand why only the I'm not sure why we announce that we don't support new-style introduction-messages when meeting a new peer. I'd think that this would be hard-coded to true, since our current code-base does support it. But maybe I'm missing something. Anyway, an obvious way to fix would be to just check the address before packing it as IPv4. |
Interesting 🤔 The assumption inside IPv8 is that old peers and old-style peers don't support anything but IPv4 and that it is, therefore, impossible for old peers to send anything to us using IPv6 addresses. However, clearly, what you say contradicts that:
I guess that still means the source is secretly IPv4, but it got remapped into the IPv6 socket (somehow? 😕), and we can safely "un-remap" it back to IPv4 when sending a response. I'm still not sure how it is possible to send to an IPv6 address from an old IPv4 endpoint, but I agree with the approach to check the address and pack it as IPv4. |
Running latest v8.0.4 on Mac.
DiscoveryCommunity and all others show 50% of overlay peers with ::ffff:192.42.1.1 like IPv6 adress. Quite a lot! 10 out of 21 peers in one overlay is IPv6.
DHT seems to fail. At least no display of them.
The text was updated successfully, but these errors were encountered: