 Caleb James DeLisle
|
caf9963793
Switch to a Rust libsodium build
|
4 years ago |
|
Caleb James DeLisle
|
98a3c41c1a
busyloop
|
4 years ago |
|
Caleb James DeLisle
|
1759ec6c07
Tune ann publish cadence based on size of messages
|
4 years ago |
|
Caleb James DeLisle
|
b3b3c3eb2a
codestyle
|
4 years ago |
|
Caleb James DeLisle
|
67d491c50a
ReachabilityAnnouncer: when it goes into MSGFULL state it never comes out
|
4 years ago |
|
Caleb James DeLisle
|
8cff3cb069
Potential null deref
|
4 years ago |
|
Caleb James DeLisle
|
3c1701ae48
Expire version and encoding scheme entities as well as peers
|
4 years ago |
|
Caleb James DeLisle
|
621d8665f5
Order of operations
|
4 years ago |
|
Caleb James DeLisle
|
67ba7d667f
Oops, computing in hours instead of minutes, also better randomizing of timeouts
|
4 years ago |
|
Caleb James DeLisle
|
c962ff2ef5
bad arithmatic
|
4 years ago |
|
Caleb James DeLisle
|
52ce0c6078
ReachabilityAnnouncer.c: Better logging
|
4 years ago |
|
Caleb James DeLisle
|
8f95a7d3c8
fat finger mistakes
|
4 years ago |
|
Caleb James DeLisle
|
ff2708f791
Switch to re-announcing peers at random times within a 5 minute range up to 1 minute before the timeout
|
4 years ago |
|
Caleb James DeLisle
|
dc9b14238b
Rewrite ReachabilityAnnouncer to not have it's own state anymore - no more risk of it desyncing from ReachabilityCollector, many bug fixes
|
4 years ago |
|
Caleb James DeLisle
|
d142d78890
Fix incorrect logic in ReachabilityAnnouncer which can cause nodes to 'disappear' from the list
|
4 years ago |
|
Caleb James DeLisle
|
784f83b4d8
Fix subnode to allow multiple peers to be announced up
|
4 years ago |
|
Caleb James DeLisle
|
4cb93d8928
Fix issue causing constant desync with route server
|
4 years ago |
|
Caleb James DeLisle
|
6c1751d868
Got rid of usages of Except and simplified Message.h
|
4 years ago |
|
Caleb James DeLisle
|
25a47b9f3c
message overflow when reporting many peers
|
6 years ago |
|
Caleb James DeLisle
|
b97787073b
BIG endian
|
6 years ago |
|
Caleb James DeLisle
|
a86c3e147b
If we reset session w/ snode, we should re-submit the whole known LinkState history
|
6 years ago |
|
Caleb James DeLisle
|
ac9e0c0022
Publish link state info to the snode
|
6 years ago |
|
Caleb James DeLisle
|
f187db6b66
Major refactoring to push data up from the InterfaceController to the ReachabilityCollector for submitting lag, drops and passed bits to the snode
|
6 years ago |
 Mada
|
a486ee7404
Handle result of updatePeer in any case
|
6 years ago |
|
Mada
|
e3101f9bd2
msgOnWire should be clear to prevent it has effact on reset state
|
6 years ago |
|
Mada
|
0cba7c15c8
Update any peer with latest local state in the timeout callback
|
6 years ago |
 Valentin Lorentz
|
98c704ce1c
Replace address checks by function calls.
|
7 years ago |
|
Mada
|
0ee9cb7f45
Reduce snode path update traffic, reset candidate when snode unreachable
|
7 years ago |
|
Caleb James DeLisle
|
02b3f25521
better assertions to catch garbage
|
7 years ago |
|
Caleb James DeLisle
|
2891fc6f8f
Actually I was right the first time, we want to send periodic empty messages which are accepted but discarded by the snode
|
7 years ago |
