summaryrefslogtreecommitdiff
path: root/node
AgeCommit message (Collapse)Author
2014-11-21Send multicasts in random order.Adam Ierymenko
This should not affect most users, but on large networks it should cause service announcements to work a lot better. This is the result of a prolonged discussion with a user about the visibility of game servers on a large network. The old multicast algorithm was de-facto randomized due to its distributed nature, while the new algorithm is more deterministic. This will restore some randomization beyond limit-overflow conditions. It won't affect small networks at all.
2014-11-20Tweak to symmetric NAT buster to add one to the number of ports above the ↵Adam Ierymenko
current one it attempts.
2014-11-20Change path selection logic to exclude non-fixed and non-active paths -- ↵Adam Ierymenko
possible fix for "NAT traversal coma" issue. Also fix a typo.
2014-11-13Optimization: we don't need to verify signatures on certs if they're certs ↵Adam Ierymenko
we already have and have verified.
2014-11-13Prevent "software laser" in legacy multicast support. Already hotpatched in ↵Adam Ierymenko
supernodes.
2014-11-04Make selftest build and run on FreeBSD 10 with gmakeAdam Ierymenko
2014-10-29Back off default multicast limit just a little.Adam Ierymenko
2014-10-29More cleanup, and fix a bug in Multicaster::gather()Adam Ierymenko
2014-10-29Moderate efficiency improvement on multicast gather result parsing, and go ↵Adam Ierymenko
ahead and keep track of total known peers.
2014-10-29Code cleanup, and fix some unsafe pointer handling in Network.Adam Ierymenko
2014-10-28Fix a nasty bug introduced in packet fragmentation a while back during ↵Adam Ierymenko
refactoring, and a few other things related to multicast.
2014-10-28Make multicast gather slightly more aggressive, and add total to list ↵Adam Ierymenko
command in testnet.
2014-10-28NULL dereference on still-initializing node bug fix in status query ↵Adam Ierymenko
commands, and doc updates.
2014-10-27Testnet seems to work a bit better now...Adam Ierymenko
2014-10-27...Adam Ierymenko
2014-10-24More testnet work.Adam Ierymenko
2014-10-23Testnet work... getting there!Adam Ierymenko
2014-10-22Some work on simulated net...Adam Ierymenko
2014-10-21Starting on simulated socket manager.Adam Ierymenko
2014-10-21More work on abstracting socket manager.Adam Ierymenko
2014-10-21About halfway there in refactoring to support pluggable SocketManager.Adam Ierymenko
2014-10-21Move socket manager implementation into osnet/Adam Ierymenko
2014-10-21Simplify locking semantics some more to address a deadlock.Adam Ierymenko
2014-10-19Don't need to check peer version for every item in txQueue since there is ↵Adam Ierymenko
only one peer.
2014-10-19Now for some clang compile fixes...Adam Ierymenko
2014-10-19Windows build fixes.Adam Ierymenko
2014-10-14Make several changes to eliminate potential deadlock or recursive lock ↵Adam Ierymenko
conditions, and add back rescan of multicast groups on network startup.
2014-10-14Get rid of rescanMulticastGroups() in Network thread since this can ↵Adam Ierymenko
deadlock... the fact that this can happen is probably bad design.
2014-10-14Fix sleeper bug that probably did nothing in setSupernodes().Adam Ierymenko
2014-10-14More fixes to legacy support, and to a potential issue on quit.Adam Ierymenko
2014-10-13Permanently retire peers.persist, but make iddb.d always enabled instead ↵Adam Ierymenko
since identities are what we really want to cache.
2014-10-13Another build fix.Adam Ierymenko
2014-10-13GitHub issue #111Adam Ierymenko
2014-10-13Yank potentially costly support for TTL out of UdpSocket since we dont use ↵Adam Ierymenko
this anymore.
2014-10-12Make crypto and compression build optimized in debug, and also try disabling ↵Adam Ierymenko
peers.persist -- might ship this way as it seems more trouble than its worth.
2014-10-11Delete bunch of commented out code.Adam Ierymenko
2014-10-11Stop persisting last announcement time since Multicaster is volatile. Also ↵Adam Ierymenko
some more legacy multicast fixes.
2014-10-11Fixes to legacy peer support.Adam Ierymenko
2014-10-10Add a sanity limit to legacy multicast repeater function in supernode-mode ↵Adam Ierymenko
nodes, and change netconf-master to issue multicast limit (ml) instead of old p5 stuff.
2014-10-10(1) Back off a little on default max multicast limit since 128 is pretty ↵Adam Ierymenko
bandwidth heavy, (2) add a little to default multicast rate limit since new MC algo is fairerererer, (3) decided not to involve netconf masters in multicast so take that out of list of who gets LIKEs.
2014-10-10<= MTUAdam Ierymenko
2014-10-09Return self in GATHER requests if self is a member of multicast group, and ↵Adam Ierymenko
reinstate legacy support.
2014-10-09New multicast bug fixes, TRACE improvements, and temporarily disable legacy ↵Adam Ierymenko
multicast for debugging purposes.
2014-10-09Reorg multicast packet, and a whole bunch of refactoring around the pushing ↵Adam Ierymenko
of certificates of membership.
2014-10-09Supernode propagation of legacy multicast frames was brokezored.Adam Ierymenko
2014-10-07Add IPv4 address assignment to ffffffffffffffff test network.Adam Ierymenko
2014-10-06Bug fix in new multicast frame handler, handling of old "P5" multicast ↵Adam Ierymenko
frames in new way.
2014-10-05Bump version to 1.0.0, add legacy support code to Multicaster to not send ↵Adam Ierymenko
new frame to known-to-be-old peers.
2014-10-04Send a VERB_P5_MULTICAST_FRAME to the supernode for each multicast -- for ↵Adam Ierymenko
now. This will allow supernodes to echo these down to pre-1.0.0 peers.
2014-10-04Pull logic to always send new multicasts to supernode since we need to do ↵Adam Ierymenko
that differently, re-add support for active bridges, and remove some gratuitous use of std::set where not needed.
generated by cgit v1.2.3 (git 2.39.1) at 2026-04-19 00:01:17 +0000