@eyedeekay
+R4SAS
+RN
+RN_
+Xeha
+not_bob
+orignal
FreeRider
Irc2PGuest75862
Onn4l7h
Onn4|7h
T3s|4_
aargh3
acetone_
anon4
cancername
eyedeekay_bnc
profetikla
shiver_1
u5657
weko_
x74a6
orignal
i2pd doesn't send ping packets
orignal
pongs only
orignal
about this RI
orignal
do you know if it was compressed or not?
zzz
not a real ping, just an empty packet with padding only we use for keepalive for introducers. Since it's empty it needs 5 bytes padding minimum
orignal
yes I send such packets
orignal
size_t payloadSize = CreatePaddingBlock (payload, 20, 5);
orignal
5 bytes min
zzz
anyway, that's what I fixed
zzz
re: RI, I've seen the issue on both compressed and uncompressed, in the RI in the session confirmed
zzz
maybe an off-by-one issue, missing the last ';' ?
orignal
was it frgamented?
orignal
I will check
orignal
I'm wornding if it was close tomax packet size
orignal
maybe buffer was full
orignal
I will check
zzz
looking...
zzz
saw several times on Sept. 15, nothing since then
orignal
that's why I think it might be a perfect match of size
zzz
I've seen it both in session confirmed (always unfragmented) and in data phase
zzz
both compressed and uncompressed
orignal
packet size matters
orignal
but I will check
zzz
packet sizes, not including IP/UDP overhead:
zzz
913, 915, 918, 991, 996, 1000, 1004, 1235
orignal
so it's not this
orignal
I will check
zzz
working on a performance section of the blog post. What was your i2pd-i2pd SSU2 vs. SSU1 max speed improvement?
orignal
R4SAS seen 1.7 Meg/s with SSU2
orignal
SSU1 is not more than 500K
zzz
ok thanks
orignal
everybody is busy in politics now ))
orignal
wait for second
orignal
how about signature?
zzz
I parse the RI before validating the sig, so I don't know
orignal
why do you do it?
orignal
anyway it's not just 1 bytes off because signature
zzz
that's the way it's always been, have to do both, doesn't matter what order
orignal
if signature doesn't match you don't waste resources to parse
zzz
sure, but if there's parse errors or invalid values, you don't waste resources validating signature :)
zzz
99.9% of the time doesn't matter, because both are successful
orignal
do you see this issue with one router or few?
zzz
looks like only one router, 176.102.64.242 / 2a02:25b0:aaaa:aaaa:6407:e5dc:1aa9:1 port 23154 dBl1ZP
orignal
maybe memory corruption
zzz
didn't realize it was all the same router until now because it was both ipv4 and v6