Adam Ierymenko
|
994565bdeb
|
Fix for GitHub issue #71
|
12 years ago |
|
Adam Ierymenko
|
eae130467b
|
Build fix for Windows setsockopt().
|
12 years ago |
Adam Ierymenko
|
f764cf8d31
|
Make UDP send and receive buffers as big as possible to reduce packet loss.
|
12 years ago |
|
Adam Ierymenko
|
4b773b61f5
|
Netconf fixes.
|
12 years ago |
|
Adam Ierymenko
|
657f6ae342
|
Don't transmit broadcasts if enableBroadcast is false on a network.
|
12 years ago |
|
Adam Ierymenko
|
2861229558
|
Windows build fixes for MAC changes (in tap driver), some comments and cleanup in MAC.hpp.
|
12 years ago |
|
Adam Ierymenko
|
66a38a9e7c
|
Default for private should be true.
|
12 years ago |
|
Adam Ierymenko
|
31ddc49da2
|
GitHub issue #67
|
12 years ago |
|
Adam Ierymenko
|
317995b921
|
Fixed for Linux tap for GitHub Issue #69
|
12 years ago |
|
Adam Ierymenko
|
beb7b5bbe5
|
GitHub Issue #69 - make MAC assignment schema differ between virtual networks.
|
12 years ago |
|
Adam Ierymenko
|
596e5dd583
|
Another sanity check on filename for GitHub issue #72
|
12 years ago |
|
Adam Ierymenko
|
b958a2d30c
|
Redis schema updates and fix for GitHub issue #72
|
12 years ago |
|
Adam Ierymenko
|
ecbcc9eb2c
|
TESTNET supernodes need different addresses.
|
12 years ago |
|
Adam Ierymenko
|
42f6aae259
|
More netconf work... set ZT_HOME and chdir there on service launch.
|
12 years ago |
|
Adam Ierymenko
|
de485ad900
|
Add testnet ports and ZT_USE_TESTNET define for contacting the test.zerotier.com net instead of the live net.
|
12 years ago |
|
Adam Ierymenko
|
99c5fae9da
|
Make Service communicate via empty-line-delimited Dictionary objects instead of the old size prefix way.
|
12 years ago |
|
Adam Ierymenko
|
1e6475fad6
|
Clean out unused netconf fields, rename a few, work on new netconf server.
|
12 years ago |
|
Adam Ierymenko
|
7831c4bfef
|
Cleanup, dead code removal, some pretty insignificant security stuff that's based on recommendations.
|
12 years ago |
|
Adam Ierymenko
|
d187ec82d4
|
Adding supernode yig.zerotier.com (Sydney, Australia) and removing mi-go.zerotier.com (Singapore) due to real-world usage data.
|
12 years ago |
|
Adam Ierymenko
|
aee742e767
|
More toward GitHub issue #56
|
12 years ago |
|
Adam Ierymenko
|
c9294c1a78
|
Prevent recursive transit of ZeroTier packets, toward GitHub issue #56
|
12 years ago |
|
Adam Ierymenko
|
b117ff5435
|
Probable fix for GitHub issue #63 - do not unite() if either path is TCP, since doing so can result in asymmetric failed NAT-t over UDP if one side has a firewall that permits outgoing UDP but not incoming.
|
12 years ago |
|
Adam Ierymenko
|
119ef5ecbf
|
More logic cleanup and some documentation / comment improvements.
|
12 years ago |
|
Adam Ierymenko
|
8fb442d81a
|
Yet more cleanup to TCP logic, this time adding a master switch and adding UDP preference in send().
|
12 years ago |
|
Adam Ierymenko
|
73153b89b4
|
Some cleanup, and use best (not first) UDP addresses for NAT-t VERB_RENDEZVOUS computation.
|
12 years ago |
|
Adam Ierymenko
|
d9836adbf6
|
.... aaaaaand... GitHub issue #61 was caused by the fact that we were no longer deleting taps on Windows!
|
12 years ago |
|
Adam Ierymenko
|
420edf23ad
|
Was not closing _shutdownSignalPipe in UnixEthernetTap either...
|
12 years ago |
|
Adam Ierymenko
|
6c5f6feb5e
|
Set close-on-exec in attempt to head off GitHub issue #61
|
12 years ago |
|
Adam Ierymenko
|
a8c12369fd
|
More tweaks to TCP logic for GitHub issue #60
|
12 years ago |
|
Adam Ierymenko
|
28a6d328a5
|
Some adjustments to TCP logic for GitHub issue #60
|
12 years ago |
|
Adam Ierymenko
|
48a1799f49
|
More Windows tap cleanup... seems solid. We'll see.
|
12 years ago |
|
Adam Ierymenko
|
bf24de43fe
|
Windows tap: be REAL REAL REAL PARANOID. Wake up sheeple.
|
12 years ago |
|
Adam Ierymenko
|
5abfb11813
|
Some installer stuff, complete refactoring of Windows side of newly split tap driver. Seems to work. Now to see if the cleanup we did here gets rid of the zombie tap device issue on Windows.
|
12 years ago |
|
Adam Ierymenko
|
0b8d6c7f4a
|
Builds and runs on Unix with EthernetTap changes, now for Windows... and for what we did this for: a refactor of the Windows tap connector.
|
12 years ago |
|
Adam Ierymenko
|
6c60305a96
|
Split EthernetTap into subclasses, work in progress...
|
12 years ago |
|
Adam Ierymenko
|
fe85426df6
|
A few more tweaks to TCP failover... seems to be switching back and forth pretty well now!
|
12 years ago |
|
Adam Ierymenko
|
158002d2d1
|
Tweak some timings and stuff.
|
12 years ago |
|
Adam Ierymenko
|
c96d3ebf8c
|
Such ping logic. So edge case.
|
12 years ago |
|
Adam Ierymenko
|
b1088a6bd7
|
Make Unix domain sockets mode 0777 so that properly authorized non-privileged users can access them (if they have auth token).
|
12 years ago |
|
Adam Ierymenko
|
f0223490be
|
Make GUI work with new control client API, make control client look for user authtoken.secret if system unreadable.
|
12 years ago |
|
Adam Ierymenko
|
81e5690410
|
More tweaks to TCP failover logic. Such edge case.
|
12 years ago |
|
Adam Ierymenko
|
700a450806
|
More tweaks to algorithm for determining when to fail over to TCP, and stop supernodes from resynchronizing unless explicitly ordered.
|
12 years ago |
|
Adam Ierymenko
|
0e1fc06a6f
|
The remove paths on send fail thing in Peer.cpp was not well thought out, and there is no point in mallocing the TCP write buffer.
|
12 years ago |
|
Adam Ierymenko
|
c3b41c289d
|
Resynchronize on startup.
|
12 years ago |
|
Adam Ierymenko
|
6ab95e0391
|
Get : out of Path.toString() since this will conflict with : in IPv6 IPs.
|
12 years ago |
|
Adam Ierymenko
|
316e8d1939
|
Build fix.
|
12 years ago |
|
Adam Ierymenko
|
f13493edb2
|
Oops... turns out we need to differentiate incoming from outgoing TCP and indeed learn incoming TCP paths. Otherwise the recipient of a TCP connection does not know to reply via TCP! Heh.
|
12 years ago |
|
Adam Ierymenko
|
595b386afc
|
Increase maximum packet fragments to make room for huge multicasts with certificates of membership tagged onto them. This likely will not actually increase the number of fragments in practice much, since these are then subsequently compressed. But the buffer needs to be allocated with room for them, otherwise these packets get dropped with a range_error exception internally.
|
12 years ago |
|
Adam Ierymenko
|
8e587ae481
|
Clean dead paths from peers.
|
12 years ago |
|
Adam Ierymenko
|
acb056e3b1
|
Small readability fix.
|
12 years ago |
