82aa3f59d6
Fix bug in ECHO handling (OK was invalid!), and use ECHO on newer peers for path confirmation. Also get rid of path confirmation circuit breaker since this causes issues with some peers and should be done more intelligently anyway.
2015-12-15 10:30:40 -08:00
94f4316a0e
Fix for possible high CPU usage on multicast queries.
2015-11-09 14:54:05 -08:00
57b71bfff0
Cluster simplification and refactor work in progress...
2015-11-08 13:57:02 -08:00
5f39d5b7ea
Further pare down Cluster messaging and rename some stuff.
2015-11-06 14:37:17 -08:00
8ef4edebbf
Deferred decode for HELLO to prevent HELLOcalypse under high load of new peers.
2015-11-05 12:22:58 -08:00
f1b6427e63
Decided to make this 1.1.0 (semantic versioning increment is warranted), and add a legacy hack for older clients working with clusters.
2015-11-02 09:32:56 -08:00
938d0a970b
TRACE build fixes.
2015-10-28 10:01:32 -07:00
da93712846
Clean up PUSH_DIRECT_PATH limits a bit more and make them a bit smarter.
2015-10-28 09:11:30 -07:00
cdc99bfee1
Add a circuit breaker for VERB_PUSH_DIRECT_PATHS.
2015-10-27 18:18:26 -07:00
88b100e5d0
More cleanup.
2015-10-27 17:59:17 -07:00
16bc3e0398
Factor out RemotePath subclass of Path -- no longer needed, just cruft.
2015-10-27 15:00:16 -07:00
a1a0ee4edb
Fix infinite loop in Cluster, clean up some stuff elsewhere, and back out rate limiting in PUSH_DIRECT_PATHS for now (but we will do something else to mitigate amplification attacks)
2015-10-27 12:01:00 -07:00
54a99d8e32
Well that was broken.
2015-10-27 11:14:07 -07:00
700c3166b7
Fix inverted sense bug.
2015-10-27 10:51:11 -07:00
9617208e40
Some cleanup, and use VERB_PUSH_DIRECT_PATHS to redirect newer peers.
2015-10-27 09:53:43 -07:00
2258e36a59
Move replication of COMs to avoid race condition.
2015-10-20 16:34:21 -07:00
59e1444b27
Finish wiring up Cluster, fix some issues with other recent changes.
2015-10-20 16:31:41 -07:00
eb79d4a2f3
Wire up peer announcement in cluster.
2015-10-20 16:24:21 -07:00
57e29857cf
Cluster work -- integrating with the rest of the code.
2015-10-20 15:27:53 -07:00
3adb183c5f
Fix bad COM attachment bug and eliminate an unnecessary redundant check.
2015-10-19 13:38:27 -07:00
95953b48f9
Do not allow VERB_RENDEZVOUS from non-upstream peers to block potential DOS vector.
2015-10-19 12:56:29 -07:00
5ce3aac929
Add rate limit on receive of DIRECT_PATH_PUSH to prevent DOS exploitation.
2015-10-16 10:28:09 -07:00
2debde3451
GitHub issue #235 , and I also see no reason not to communicate with people from other Worlds.
2015-10-15 07:22:17 -07:00
c312ae221f
Fix for world size in OK(HELLO)
2015-10-14 10:45:33 -07:00
123c466843
Full integration of World and World updates.
2015-10-13 12:17:47 -07:00
5d2f523e81
World stuff...
2015-10-13 12:10:44 -07:00
cae58f43f1
More World stuff, and mkworld.
2015-10-13 08:49:36 -07:00
1b1945c63e
Work in progress on refactoring root-topology into World and adding in-band updates.
2015-10-12 18:25:29 -07:00
eff1fe3c61
Create files for each hop (more convenient) and fix a packet parse bug.
2015-10-09 16:22:34 -07:00
aec13b50fd
Be a bit more verbose in circuit test reports to more clearly track current and upstream hop in graph traversal history.
2015-10-09 15:05:26 -07:00
0c498556d5
Unroll Salsa20 fully for a little more speed (non-SSE now almost as fast as SSE)
2015-10-09 09:39:27 -07:00
a3876353ca
Abiltiy to post a test via the controller web API, and parsing of CIRCUIT_TEST_REPORT messages.
2015-10-08 13:25:38 -07:00
0ce0bc00d2
Make sure received() gets called for some new messages, and docs.
2015-10-07 16:20:54 -07:00
69b44bf9a5
Finally add an ECHO.
2015-10-07 16:11:50 -07:00
73cafbe0ec
Limit proof of work difficulty to something sane.
2015-10-07 13:46:44 -07:00
e5f168f599
Add proof of work request for future DDOS mitigation use.
2015-10-07 13:35:46 -07:00
13f14c2f4c
Kill debug line.
2015-10-07 10:56:47 -07:00
ab0228f626
More cleanup and simple refactoring, consolidate InetAddres serialize/deserialize into the class.
2015-10-07 10:30:47 -07:00
1b2cac0cc5
Trim some cruft that is not used and probably never would be.
2015-10-07 09:38:33 -07:00
3593fb3462
Send initial CIRCUIT_TEST packet.
2015-10-06 15:16:41 -07:00
d3f29d09e8
Plumbing through circuit test stuff.
2015-10-06 14:42:51 -07:00
5341afcdcd
Handling of CIRCUIT_TEST, should be ready to test.
2015-10-06 11:47:16 -07:00
a3db7d0728
Refactor: move network COMs out of Network and into Peer in prep for tightening up multicast lookup and other things.
2015-10-01 11:11:52 -07:00
1a4f16e0ed
More work on circuit testing...
2015-09-30 13:59:05 -07:00
f69454ec98
(1) Make ZT_ naming convention consistent (get rid of ZT1_), (2) Make local interface a full sockaddr_storage instead of an int identifier, which turns out to be better for multi-homing and other uses.
2015-09-24 16:21:36 -07:00
367ffde00c
Plumb through localInterfaceId to track local interfaces corresponding with remote addresses.
2015-09-23 13:49:56 -07:00
0d386f1c31
Add a bit of useful testing instrumentation to SqliteNetworkController.
2015-09-08 11:35:55 -07:00
b31071463c
Try another NAT traversal improvement.
2015-07-28 11:28:47 -07:00
b3516c599b
Add a rate limiting circuit breaker to the network controller to prevent flooding attacks and race conditions.
2015-07-23 10:10:17 -07:00
3ba54c7e35
Eliminate some poorly thought out optimizations from the netconf/controller interaction,
...
and go ahead and bump version to 1.0.4.
For a while in 1.0.3 -dev I was trying to optimize out repeated network controller
requests by using a ratcheting mechanism. If the client received a network config
that was indeed different from the one it had, it would respond by instantlly
requesting it again.
Not sure what I was thinking. It's fundamentally unsafe to respond to a message
with another message of the same type -- it risks a race condition. In this case
that's exactly what could happen.
It just isn't worth the added complexity to avoid a tiny, tiny amount of network
overhead, so I've taken this whole path out.
A few extra bytes every two minutes isn't worth fretting about, but as I recall
the reason for this optimization was to save CPU on the controller. This can be
achieved by just caching responses in memory *there* and serving those same
responses back out if they haven't changed.
I think I developed that 'ratcheting' stuff before I went full time on this. It's
hard to develop stuff like this without hours of sustained focus.
2015-07-23 09:50:10 -07:00
