IRCaBot 2.1.0
GPLv3 © acetone, 2021-2022
#ls2
/2022/07/06
orignal but it contains caps, s and i only
orignal but endpoiunt comes with message
zzz right. I do that. If that fails, then I send 65
orignal so they right behaviour to let 5 go through
orignal even if Alice doesn't expect incoming connections because unpublised
zzz sure. otherwise firewalled routers couldn't do peer test at all
orignal agree
orignal will fix that part
zzz just make sure the caps has 4 or 6 or 46 if there's no host
orignal yes, it's always there
zzz I also found a SSU 1 bug where I wasn't putting MTU in firewalled/hidden addresses sometimes
orignal almost forgot about mtu
orignal need to implement it
orignal checked my code I'm fine with unpublished address
orignal I just set missing endpoint
orignal but 70 is another story
orignal lol. I send address in 6 and 7
RN send all the addresses! LOL
RN just kidding
orignal how do we use address block in msg 6?
zzz right now I ignore it
orignal maybe we shouldn't send it at all
orignal in my opinion it's another risk
orignal an advesary can initiate a peer test to you
orignal then send worng address in 6 and confuse you and let you believe you are behind a symmetric NAT
orignal unlike other messages 6 is unsolicited
zzz sure but charlie could also be an adversary for alice
zzz if charlie does trust alice, it's a "free" peer test for charlie
orignal but Alice is in peer testing stage
orignal while Charlie is not
zzz yup
orignal Alice initaates peer test, even if Charlie send crap in 7 Alice will detect symmetric NAT and try somebody else
zzz I wonder why I said it's required in 6...
zzz doesn't hurt anything though
orignal and I menationed only 7 )
orignal if I'm an advesary I can easily bring whole network to "Symmetric NAT" state
orignal if address block is handled in 6
zzz hmm
orignal <R4SAS> 18:07:16@781/debug - SSU2: PeerTest msg=4 code=69
orignal where code 69 comes from and why>?
zzz public static final int TEST_REJECT_CHARLIE_BANNED = 69;
orignal yes I know what it means
orignal but why his router was banned?
orignal i2pd never sends 69
zzz there's 10 or 20 different reasons why we could ban somebody.
zzz if my router sent the 69, I can look
orignal but what your code does?
zzz if the router is banned, it sends a 69 :)
zzz that's the whole reason we're sending hashes in the requests. so charlie can decide whether he wants to talk to alice or not
zzz that's literally why we designed it this way
orignal that's our question what's wrong with his router
zzz did my router send the 69?
orignal we don't know who was Charlie
orignal 69 came from Bob
orignal good question
orignal I should print Charlie's hash in case of error
orignal since Bob sends Charlie's hash
zzz is he BpAT?
orignal gpUB is I remember
zzz here's one where I sent a 69. BpAT is alice; gpUB is bob
orignal and what was the reason for 69 in this case?
zzz 07-06 20:49:39.703 DEBUG [ handler 1/1] .transport.udp.PeerTestManager: Got peer test msg: 2 status: 0 hash: [Hash: BpATV4o2-r6~pbHjtleL5Qgo7Wn-5f1N6oGROvxyiEg=] nonce: 1505935422 time: Jul 6, 2022, 8:49 PM ip/port: [2001:470:1f12:e56:0:0:0:2]:24768 from [2001:470:1f13:e56:bee:0:0:1]:17981 gpUBQf OB2 recvAge: 11s sendAge: 11s sendAttemptAge: 12s sendACKAge: 2013ms lifetime: 10m RTT: 17 RTO: 1000 MTU: 1280 LMTU: 1500
zzz cwin: 92560 acwin: 92560 SST: 524288 FRTX? false consecFail: 0 msgs rcvd: 45 msgs sent: 79 pkts rcvd OK/Dup: 125/0 pkts sent OK/Dup: 79/0 IBM: 0 OBQ: 0 OBL: 0 theyRelayToUsAs: 165483802 state: null
zzz 07-06 20:49:39.703 DEBUG [ handler 1/1] .transport.udp.PeerTestManager: Send msg 3 response 69 nonce 1505935422 to [2001:470:1f13:e56:bee:0:0:1]:17981 gpUBQf OB2 recvAge: 11s sendAge: 11s sendAttemptAge: 12s sendACKAge: 0ms lifetime: 10m RTT: 17 RTO: 1000 MTU: 1280 LMTU: 1500 cwin: 92560 acwin: 92560 SST: 524288 FRTX? false consecFail: 0 msgs rcvd: 45 msgs sent: 79 pkts rcvd OK/Dup: 125/0 pkts sent OK/Dup: 79/0 IBM: 0
zzz OBQ: 0 OBL: 0 theyRelayToUsAs: 165483802
zzz I don't know. it appears it was temporary
zzz I see a lot of errors with BpAT - retransmissions, bad conn ids, ...
orignal strange
zzz I don't think you've fixed all your ack problems yet. All my failures where I never get an ack back after 4 retransmissions are from i2pd
zzz BpAT, kyY2, ImQC, 2RRY, CDFn, bAU~, xZ9n, ...
zzz we have to focus on the basics. sending messages and sending acks. peer test and relay can wait
orignal let me check
orignal do you have an example of Ack with missing messages