orignal
no, enabling NTP is bad idea
orignal
we discussed it before
zzz
1-hop test results: of routers with > 3 test fails in 24 hours, 49% i2pd, 51% java, 23% RU
zlatinb
what about router caps and versions?
zlatinb
I predict Xfr and 0.9.52 for the i2pds
zlatinb
and if you feel like tracerouting them I predict they will be vps-es in ru
zzz
didn't keep track, but almost ll 52
zzz
*all
zzz
mostly non-ff
zzz
1.3 % of the known routers had > 3 fails
zzz
26% of routers in netdb are i2pd, so i2pd about twice as likely to fail
zlatinb
what % are RU?
zzz
23% RU
zlatinb
so no deviation there
zlatinb
you mean 23% in the netdb and 23% of the >3 failures?
zzz
the latter only
zlatinb
and what is it of the netdb?
zzz
maybe 10-20%, I know how many RU I have but not how many total with a successful geoip
zlatinb
are you using STBM or VTBM?
zzz
both
zlatinb
setting router.excludePeerCaps=KLMNOP so that only X is used increased the likelihood of hitting a failing peer in my test