Star signaling servers down

Hi guys,

anyone knows what happened to the following singaling servers?
https://wrtc-star.discovery.libp2p.io/
https://star-signal.cloud.ipfs.team/

This one seems to be up, but I can’t get any information about connecting peers:
https://ws-star.discovery.libp2p.io/

Anyone? Thanks!

@mtojek I am looking into the webrtc signaling server issue (fyi those domains are aliases). I’ll let you know when I have more info.

For the websocket star server can you clarify what you mean, I’m not sure I understand? Are you unable to discover and connect to peers that are listening on that server?

Oh, didn’t notice that they’re aliases - thanks for bringing this up and looking into the issue. Kindly please to keep us informed.

Regarding the websocket star server - I can see the issue only while using the server https://ws-star.discovery.libp2p.io/ . The only message I receive from the endpoint is the initial one. I don’t get any “ws-peer” updates.

Can you share your config for libp2p? I was able to get peer events, my assumption is that you need to pass ws.discovery to the peerDiscovery config, otherwise you won’t get any peer:discovery events.

Can you translate this into protocol commands? What I see, is:

time ws “wss://ws-star.discovery.libp2p.io/socket.io/?EIO=3&transport=websocket”
< 0{“sid”:“GHBd2YlYZE6RJ10OALv8”,“upgrades”:[],“pingInterval”:25000,“pingTimeout”:5000}
< 40
> 42[“ss-join”,"/hello-world"]

no output messages received. The session remains alive if I send:

42[“ping”]

That is part of the socket.io spec, which websocket-star is currently using. The most recent version I could find is at https://github.com/socketio/engine.io-protocol.

I believe 42 indicates message (4) and ping (2), which means the client has 5 seconds, based on the initial timeout returned, to ping or the connection will be closed.

I know all these message types. Just wanted to highlight that in case of "ws-star.discovery.libp2p.io " there are no incoming updates from the star signal server if I send only “ss-join”. The other server pushed updates without any issues.

So the question is: which commands should I send to receive “ws-peer” from the signal server?

Apart from that, do you know what’s going on with the mentioned signal servers? They’re still down. Any ETA to fix it?

EDIT:

I think I see the difference. “ws-star.discovery.libp2p.io” requires a crypto challenge (CRYPTO_PUZZLE.md). The other one didn’t. Is there an option to skip it?

https://wrtc-star.discovery.libp2p.io is back up. We’re going to be working on upgrading the deployment of this in the coming weeks to help prevent prolonged downtime.

Thanks for responding.

Do you guys have plans for enabling crypto challenge on wrtc-star.discovery.libp2p.io as well or this one will stay as is (no signatures required)?

I don’t anticipate us making any significant changes to the signaling server. Any updates will likely just be code/dependency, and deployment updates. Our focus right now is to get distributed signaling in place, https://github.com/libp2p/specs/pull/159, so that we can migrate away from using these centralized services.

Ok, thank you for all replies, will watch the distributed signaling thread.