jrandom
c2d55013a6
0.3.4.2, backwards compatible, release pretty soon
2004-08-13 02:15:54 +00:00
jrandom
25eda1378e
* do DoS detection in constructor, so we get useful "why are we doing this"
...
stack traces (rather than "oh, we're doing it when... uh... writing to the socket")
* increase the throttle max, since we want to be able to send a few concurrent
2004-08-13 02:11:54 +00:00
jrandom
dfac7bde9c
* track searches more carefully
...
* detect situations where we may be inadvertantly flooding the netDb
and log them as CRIT with a stacktrace, as well as publish the count
of those events in the netDb
* detect potential netDb DoS situations by checking to see if we have
received more than 20 netDb lookups in 10 seconds, and if so,
probabalistically drop subsequent netDb messages (P=1-(10/numReceived)).
This is also published in the netDb.
2004-08-13 01:43:01 +00:00
jrandom
348168d6c0
made the log less verbose for duck ;)
2004-08-13 01:27:55 +00:00
hypercubus
a1c772c8d8
changed quadn.i2p to library.i2p
2004-08-12 21:02:43 +00:00
jrandom
f1ce1b5361
if we reach this point, bump up the expiration if necessary (otherwise the fast expire will occur with small clock skews)
2004-08-12 03:24:44 +00:00
jrandom
ebdc7d70a1
shitlist appropriately (continued)
2004-08-12 03:23:48 +00:00
jrandom
c5947c23bb
include the shitlist summary
2004-08-12 03:22:27 +00:00
jrandom
eeb1852d95
take note of the reason each peer is shitlisted and display that on the console (good idea oOo)
...
cleaned up the shitlisting process within the TCPTransport so that we don't shitlist twice (clobbering the detailed cause with a general "uh, couldn't contact 'em" cause)
2004-08-11 22:51:00 +00:00
jrandom
2f28a635a9
integrated oOo's memory usage patch (the % shown is unfortunately % of peak, not % of max) [thanks oOo!]
...
cleaned up build script to make sure we always build fresh JSPs and xml [thanks oOo^2!]
2004-08-11 22:23:48 +00:00
jrandom
d524c77560
*cough* (oops)
2004-08-11 08:23:56 +00:00
jrandom
0025d94aa4
if the message has expired but hasn't exceeded the fudge factor, we still need to give it some time to be processed (we receive and validate it first, and if it reaches these jobs, we should let it continue)
...
logging
2004-08-11 07:10:37 +00:00
jrandom
bb5ae2922d
added freshcoffee.i2p
2004-08-11 07:00:23 +00:00
jrandom
fbe9fe1ba8
new method for sanitizing content to be rendered on a web page (specifically, the routerInfo options can be hacked to contain evil html)
...
thanks for finding and suggesting the fix oOo!
2004-08-11 04:42:04 +00:00
mpc
007194d674
Check mallocs for NULL return
2004-08-11 04:39:17 +00:00
jrandom
cdd74505d7
optional flag to allow using the standard base64 charset on decode/encode, rather than the (filename friendly) i2p charset
2004-08-11 00:59:11 +00:00
jrandom
0aa023189d
shitlist the old idents of peers who change idents
...
be sure to fire any onFail jobs if we aren't going to pass a message off (duh)
take clock skew into account when determining whether a message is expired (duh^2)
2004-08-11 00:57:36 +00:00
duck
79aa10dfcb
just a test
2004-08-10 21:22:45 +00:00
jrandom
9ecfda0110
added basic HTTP authentication for accessing the router console (if a consolePassword is specified in the router.config)
...
unfortunately, this password setting is only read on router startup...
2004-08-10 19:51:11 +00:00
jrandom
b89e26c460
cleaned up the validation of leaseSet/routerInfo elements being accepted so we validate only in one place (in the facade instead of both the facade and the dataStore)
...
don't accept entries created (too far) in the future
2004-08-10 16:55:54 +00:00
jrandom
97e5952544
another oOo catch (not sure if this is ever run - it shouldn't be, but now it'll display the silly statement more clearly :)
2004-08-10 04:52:48 +00:00
jrandom
8627328047
set the nextHopId whenever we know the nextHopInfo (duh) [thanks oOo!]
...
this should fix the deserialization errors from tunnels.dat
2004-08-10 04:11:58 +00:00
jrandom
ec0c912c6f
oOo's updates:
...
* properly describe expired leaseSets (e.g. "30s ago" instead of "in -30s")
* add a little table at the end of the netDb HTML summarizing the versions people claim to be running
2004-08-10 01:07:33 +00:00
mpc
953de3f1f2
Indentation
2004-08-09 10:56:14 +00:00
mpc
e1264de514
Magically fixed itself
2004-08-09 10:02:07 +00:00
sunshine
5abd2b400c
Updated wiki text sources
2004-08-09 10:01:04 +00:00
mpc
2c2a103676
Some more Makefile updates
2004-08-09 09:39:20 +00:00
hypercubus
44af799b66
reverting previous commit; realized it would hammer Mort Bay's servers
2004-08-08 23:07:07 +00:00
hypercubus
ec22ba3248
'dist' target broken, fixed by adding prepWEB as a depend before buildWEB
2004-08-08 22:42:56 +00:00
jrandom
7fcc05c037
0.3.4.1, backwards compatible, release later today, blah blah blah
2004-08-08 19:29:53 +00:00
jrandom
edf17d0a46
drop the arbitrary wait during creation (but not during startup)
...
only relevent to simulations, obviously
2004-08-08 19:26:44 +00:00
jrandom
9cccd0bfc9
rather than flush any/all log messages 10 times a second, flush log messages once there are 100 of them or 10 seconds have passed, whichever comes first
2004-08-08 01:40:48 +00:00
jrandom
e57c010e3d
reduced context switches by removing the timeout val that was redundant
2004-08-08 01:39:00 +00:00
jrandom
4dfcf1c1c8
updated boundary condition (thanks mule/oOo)
2004-08-08 01:35:45 +00:00
jrandom
8d7786e97d
bad jrandom, test first
2004-08-08 01:34:42 +00:00
jrandom
2cb519cd06
updated implementation to match the RandomSource javadocs (inclusive of both 0 and n, unlike SecureRandom, which is inclusive of 0 and exclusive of n)
2004-08-07 19:14:13 +00:00
jrandom
bc46ad4331
only block adding more *outbound* data, not handling data received from I2P. The data has already been received by the router and delivered to the streaming lib (and is sitting in RAM anyway...)
...
logging
2004-08-07 06:35:46 +00:00
jrandom
be08e8f23b
use the env defaults when no i2cp info is specified
2004-08-07 06:28:29 +00:00
jrandom
f937809903
have a write() timeout after 60s instead of blocking forever by default (also used when injecting data into an input stream)
2004-08-06 22:24:56 +00:00
jrandom
c0f32c942d
more checking for timeout
2004-08-06 22:23:00 +00:00
sunshine
39c5c830bb
Added py2exe example
2004-08-06 14:41:05 +00:00
sunshine
83c8953d1b
Removed debug print statements
2004-08-06 13:43:46 +00:00
sunshine
4b100a5a64
Replaced time.clock() -> time.time() (Bug on Unix)
2004-08-06 13:43:22 +00:00
sunshine
b7e50e0b3a
Replaced time.clock() -> time.time() (Bug on Unix)
2004-08-06 13:36:41 +00:00
mpc
6933052de7
Minor improvments to the build system
2004-08-06 01:27:53 +00:00
jrandom
22d945f7b7
allow remote harvesting by specifying a URL to read routers from (e.g. --netDbURL http://dev.i2p.net/i2pdb/ )
2004-08-05 20:35:52 +00:00
jrandom
b81c5628ce
fixed harvest of rankings (capacity vs reliability thing)
2004-08-05 20:07:10 +00:00
jrandom
cdb4576bd7
per mule's patch, added support for a round robin across multiple outproxies
...
instead of the standard 'httpclient 4444' or 'httpclient 4444 squid.i2p', you can now specify a comma delimited list of outproxies:
'httpclient 4444 squid.i2p,www1.squid.i2p,www2.squid.i2p' and each individual http request goes through a randomly selected proxy
there are a few general issues with this, such as a lack of affinity (web applications that require a session to always come from a single IP address will break)
but it should work most of the time.
2004-08-05 18:17:33 +00:00
jrandom
4859cd7dcf
new job to aggressively fetch the most up to date routerInfo from peers by, er, asking them for it
...
disabled by default (enable through router.config 'netDb.shouldHarvest=true').
useful if you want to harvest the published router stats, since otherwise you could have data from a few hours back
2004-08-05 02:58:13 +00:00
jrandom
3f70593ca8
precompile all of the JSPs so that people with JREs instead of JDKs can load this up
2004-08-05 02:24:00 +00:00
