Pre-Dev-Update Thread! Yay! :D

IGD is a router standard I think. Some home routers can handle it, others can’t. Think there was something about it in the last dev update.
I’m stuck in the car waiting for mother in law to do shopping so I can only look on as you lot get to play :cry:

2 Likes

Different error gives fuller suggestion:

INFO 2020-06-11T15:09:47.823482273+01:00 [/usr/local/cargo/git/checkouts/quic-p2p-9a2715614ed5c99d/c6ae051/src/lib.rs:335] IGD request failed: Could not find the gateway device for IGD - IgdSearch(IoError(Custom { kind: TimedOut, error: “search timed out” }))
ERROR 2020-06-11T15:09:47.823614977+01:00 [src/bin/safe_vault.rs:169] Cannot start vault due to error: Routing(Network(IgdNotSupported))
ERROR 2020-06-11T15:09:47.823658178+01:00 [src/bin/safe_vault.rs:170] Automatic Port forwarding Failed. Check if UPnP is enabled in your router’s settings and try again. Note that not all routers are supported in this testnet. Visit https://forum.autonomi.community for more information. :smiley:

(I’ve been thinking I need to get a new supplier… and with that a new router)

2 Likes

I am soo confudazzled, my logs say…

  • This node has been approved to join the network at Prefix()!
    but then…
  • Could not send the IGD response: sending on a disconnected channel - “SendError(…)”

time to move from pre-dev-update to dev-update-official :woozy_face:

Note that at the time of writing vaults from home is being restricted to those with home routers which correctly implement IGD. This will be expanded imminently to include those with routers which don’t support IGD, with instructions added here for manual port forwarding at that point. If your log file states Automatic Port forwarding Failed then be on stand by for the next iteration.

2 Likes

I love how you guys are already all over this even before the update tonight :+1:
some real anticipation anticipation here!

8 Likes

@Josh We’ve seen this a couple of times as well but, so far anyway, if we’ve left the vault running it has eventually functioned normally in the section.

So to be clear for anyone reading, if you see the error

Automatic Port forwarding Failed. Check if UPnP is enabled in your router's settings and try again.    Note that not all routers are supported in this testnet. Visit https://forum.autonomi.community for more information.

then unfortunately it means your router does not correctly support IGD and therefore you’ll need to skip this particular version of the testnet as IGD is what we want to test in isolation. However if you see the error in your logs

Could not send the IGD response: sending on a disconnected channel - "SendError(..)"

then leave your vault running as our tests show this eventually resolves itself and your vault will join the testnet, graduate to Adult, etc :slight_smile:

And if you see a completely different error, do of course let us know!

11 Likes

thanks for sharing it!

3 Likes

Thanks @StephenC, I checked as I had left it running and it is working just fine! :partying_face:

Yes! thanks @davidpbrown

6 Likes

WOOT!
https://github.com/maidsafe/safe-nd/pull/163

18 Likes

Yes there has been a fair amount of github action in tasty places :slight_smile:

6 Likes

As the Sequenced data type is merged to master it’s like a damn is breaking:

10 Likes

They keep coming. Goodbye AppendOnly and hello SequenceData

https://github.com/maidsafe/safe-nd/pull/168

15 Likes

Dam tootin’ pardner…

3 Likes

I’m curious, is the bft-crdt being used in safe-transfers? Maybe I haven’t noticed it yet but haven’t seen bft-crdt added to the Maidsafe repos yet.

5 Likes

Not at this time, we specialised it for ourselves which is fine as it’s not a lot of code at all. As we get to testing we may then use the more generic bft-crdts but in any case we hope it spurs lots of people to look into this way of working.

So not at the moment, but we will see. David Rusu did a brilliant job there and I think he really enjoyed it as well.

15 Likes

Okay. When you say more generic bft-crdt is that implying that the specialized CRDT data types for SAFE are bft in nature?

4 Likes

We already have a secure broadcast mechanism and basically in each section a network of 7 nodes that decide on things. So the specialisation is mostly that. As the code is so few lines rather than generalising we use the at2 pattern in our network. So we have sections, keys and digital signatures and specialise on that.

9 Likes

Thanks for the list! @adam
https://github.com/maidsafe/routing/issues/2117

13 Likes

Wow… maybe the last piece of masterpiece…

6 Likes

Yes it clears the way for the simpler, more natural approach and should make vaults much much less resource hungry (a goal we need to stick to). It’s a fair lump of work, but well worth it.

13 Likes

https://github.com/Yoga07/safe_vault/commit/4425134afa7664471a8f47577328564ce77f1264

at2 creeping in as the devs push!

21 Likes