IRCaBot 2.1.0
GPLv3 © acetone, 2021-2022
#i2p-dev
/2022/12/15
@eyedeekay
&eche|on
&kytv
&zzz
+R4SAS
+RN
+RN_
+dr|z3d
+hk
+orignal
+postman
+weko
+wodencafe
An0nm0n
Arch
Danny
DeltaOreo
FreefallHeavens
Irc2PGuest21881
Irc2PGuest5995
Irc2PGuest88897
Irc2PGuest95630
Nausicaa
Onn4l7h
Onn4|7h
Over
Sisyphus
Sleepy
Soni
T3s|4_
T3s|4__
aargh2
acetone_
anon2
b3t4f4c3
bak83
boonst
cumlord
dr4wd3_
eyedeekay_bnc
hagen_
khb_
not_bob_afk
plap
poriori
profetikla
r3med1tz-
rapidash
shiver_
solidx66
u5657
uop23ip
w8rabbit
x74a6
eche|sleep sorry,mixed the date yesterday...
dr|z3d here it is:
dr|z3d ERROR […Establisher] …sport.GetBidsJob: Send a message to ourselves? nuh uh…
dr|z3d • Contents: OutNetMessage [Priority: 1000] with 83 byte DatabaseLookupMessage [MsgID 2652637312]
dr|z3d • Delivery failure on transports [SSU]
dr|z3d • Expires: Thu Dec 15 13:10:37 GMT 2022
dr|z3d java.lang.Exception: I did it
dr|z3d at net.i2p.router.transport.GetBidsJob.getBids(GetBidsJob.java:67)
dr|z3d at net.i2p.router.transport.CommSystemFacadeImpl.processMessage(CommSystemFacadeImpl.java:172)
dr|z3d at net.i2p.router.OutNetMessagePool.add(OutNetMessagePool.java:51)
dr|z3d at net.i2p.router.transport.TransportImpl.afterSend(TransportImpl.java:399)
dr|z3d at net.i2p.router.transport.TransportImpl.afterSend(TransportImpl.java:298)
dr|z3d at net.i2p.router.transport.udp.UDPTransport.failed(UDPTransport.java:3436)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager.establish(EstablishmentManager.java:329)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager.establish(EstablishmentManager.java:318)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager.handlePendingIntro(EstablishmentManager.java:1598)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager.handleOutbound(EstablishmentManager.java:2536)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager.access$1400(EstablishmentManager.java:68)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager$Establisher.doPass(EstablishmentManager.java:3164)
dr|z3d at net.i2p.router.transport.udp.EstablishmentManager$Establisher.run(EstablishmentManager.java:3121)
dr|z3d at java.base/java.lang.Thread.run(Thread.java:1589)
dr|z3d at net.i2p.util.I2PThread.run(I2PThread.java:103)
postman ohh i got these too aplenty
dr|z3d oh, you do? they're pretty rare, which is why we've got an "I did it exception" so zzz can trace the source of the issue.
zzz postman, what router version are you running?
zzz ok, SSU2 introducer issue. We're alice, trying to connect to charlie who has us as an introducer (bob), so we're trying to connect to ourselves
zzz didn't think about that case
zzz should be easy fix
zzz thanks dr|z3d
dr|z3d great, zzz, nice quick analysis.
zzz so unfortunately it was not something I broke recently
dr|z3d well, it's a pretty rare bug, and it doesn't look like it's hurting anything.
dr|z3d so that's something.
dr|z3d not a "looks like we need a point release" type issue.
zzz ultimately harmless, yes
y2kboy23 I'm on -7 and i2psnark is acting slow. I use to be able to connect to most of the peers, upload and download, and would average 100 kbs or so. Now, I'm connecting to about half the swarm and speeds are around 50 kbs. The most alarming was I saw on the tunnels page for i2psnark, 35GB down and 21GB up. That's not remotely close to my actual usage.
mesh y2kboy23: speed varies a lot when it comes to i2psnark
y2kboy23 I understand that. I don't understand how over stated the download/upload parts are
y2kboy23 At most it was a couple of GBs on both sides.
mesh y2kboy23: you might try simply restarting. You could also try going to I2psnark > settings and increasing the number of tunnels and/or decreasing the number of hops i2psnark uses.
mesh y2kboy23: if your problem is a low "router.activePeers" count then I think what helps is increasing exploratory tunnels. I tend to drastically max out the number of exploratory tunnels that the router builds and this helps to make sure the router is well connected I think
y2kboy23 I've restarted quite frequently with the dev builds. I know there was work done with the tunnels logic, I wonder if this is a down stream issue from that work.
y2kboy23 Ill attempt going back to the 2.0 release and compare.
mesh y2kboy23: go to /configtunnels and check how many exploratory tunnels you're using and how many client tunnels for i2psnark and i2p http proxy
y2kboy23 mesh 3/3 for snark. I just upped the explorer to 4+1/4+1
mesh y2kboy23: well I'm no expert but I tend to use many more tunnels than that. 8 tunnels, 2+1 for exploratory. 10 tunnels, 2 for i2psnark.
mesh y2kboy23: but this is partly because I'm trying to compensate for being in "Hidden" mode and partly because my computer is powerful enough to handle it. Even though i2p screams "PERFORMANCE WARNING - Settings include high tunnel quantities" I tend to notice little real impact
mesh then again I have this problem where after a few days of uptime my router sometimes seems to explode
y2kboy23 mesh mine doesn't explode. It's generally been stable and chugging along.
mesh y2kboy23: I also run i2p with the low latency gc. This actually seemed to help a lot when I looked at stats like "Job Lag" and "Message Delay" and led to better performance. Generally I get the best performance by using the latest jdk (jdk19) and -XX:+UseZGC -XX:SoftMaxHeapSize=2g -XX:+UseNUMA -XX:+UseStringDeduplication
mesh my router tends to avg ~400-500mb of memory which is double the 256mb I think. And it's memory I don't really miss since I have a bunch of ram on this devbox
mesh you wouldn't want to run so many tunnels on your phone
y2kboy23 mesh I think I've allocated 512mb to the router. I could allocate more if needed.
mesh y2kboy23: instead of allocating too much or too little, you can use ZGC and SoftMaxHeapSize. This lets the router/jvm use whatever is best. Whatever memory it doesn't need will be returned to the OS
mesh sometimes my router uses more than 512mb, especially if I'm running a lot of torrents it can get up even into the 700s
mesh y2kboy23: everything okay?
y2kboy23 mesh I am now. I was fighting the server for a bit.
zzz holy smokes
zzz I've been connected to from our #2 bad router via 80 different IPs in 40 minutes
zzz no single IP twice. no common subnets
dr|z3d that's the one I think I was having issues with on the blocklist, zzz.
dr|z3d the same one that seems to not be publishing any ips, introducers or otherwise, to the netdb.
dr|z3d 80 ips in 40 minutes is a bit special. *chuckle*
dr|z3d hostile tunnel requests, seeing a fair few of those lately. don't remember seeing (m)any before.
dr|z3d duplicate hops, stale requests from yesterday..
dr|z3d "Dropping hostile build request with the same previous and next hop..." and more.
dr|z3d router in question is also happily evading blocklist again.
dr|z3d (and still no ip or introducers in the netdb entry) .. you seeing any, zzz?
dr|z3d looks like you been busy just now.
zzz -8 will actually drop banned peers inbound in SSU 1/2, was no check before, a major omission
dr|z3d oh, splendid.
zzz Blinded message
dr|z3d very nice.
dr|z3d that's session only on the ip address, presumably?
zzz yeah
dr|z3d good work on the blocklist, now working. the first block on the session is offender #2. :)
zzz I stopped counting at 110 IPs
dr|z3d that's definitely suspect behavior. I can't think why you'd be switching ips so rapidly, even on a vpn.
dr|z3d not doing any counting here at all for 1&2. they're already permabinned.
zzz you can't permaban an inbound SSU connection that's changing IPs because you don't know the router hash until the end of the handshake. That's how I got to 110