 Caleb James DeLisle
|
4bd18f7672
If a search causes you to communicate directly with the target, ask them for their peers
|
9 years ago |
|
Caleb James DeLisle
|
258a2c4a46
it all seems to work well enough
|
9 years ago |
|
Caleb James DeLisle
|
ec8c4fdd62
Added -c to search to allow for more exhaustive searches, also sort sessionStats
|
9 years ago |
|
Caleb James DeLisle
|
ff287fbbff
getpeers should not splice the address in order to be clearer
|
10 years ago |
|
Caleb James DeLisle
|
7667736795
NodeStore_brokenLink() instead of NodeStore_brokenPath()
|
10 years ago |
 Arceliar
|
00e46c95aa
Some cleanup, mainly to the Janitor. Moved most of the identical code for getPeers-ing from different rumormills into one function. Rewrite the keyspace maintenance code to add important addresses to a rumormill so it can ping the relevant nodes (or do a search) later. Sends fewer searches but more pings. Expect about the same number of packets per minute as before for the current network, can potentially increase slightly (logarithmically up to a maximum of ~2x what we used before) as the network grows.
|
10 years ago |
|
Arceliar
|
c8d31c3a39
Waste less bandwidth on searches that are usually for nonexistant addresses anyway
|
10 years ago |
|
Arceliar
|
f4c653ccb1
Use a 3-RumorMill system.
|
10 years ago |
 Caleb James DeLisle
|
0fdda4fc72
Replaced all instances of Assert_always() with Assert_true() to standardize on Assert_true() for 'always', Assert_ifParanoid() for assertions which trigger if PARANOIA is enabled and Assert_ifTesting() for asserions which would be triggered by bad nodes in a real network
|
10 years ago |
|
Caleb James DeLisle
|
72505b0b7b
Replaced more expensive instances of Assert_true() with Assert_ifParanoid()
|
10 years ago |
|
Caleb James DeLisle
|
177cc6946f
Better checking to spot cause of assertion failure and more initial nodes in search
|
11 years ago |
|
Caleb James DeLisle
|
1f1cf83d4e
Work from over the weekend, crashey is less crashey (at least in the sim)
|
11 years ago |
|
Caleb James DeLisle
|
e569a8cb76
Tuning still needed but this one has the sound of a winner
|
11 years ago |
|
Caleb James DeLisle
|
6a142b8c91
Fixed pathologically bad MAX_REQUESTS_PER_SEARCH
|
11 years ago |
|
Caleb James DeLisle
|
094e5b94e0
Fixed a few more bugs
|
11 years ago |
|
Caleb James DeLisle
|
cca56ad9a9
Finally got everything switched over to using the new pathfinder
|
11 years ago |
|
Caleb James DeLisle
|
2c25a476b3
Moved the code over to completely use the new pathfinder
|
11 years ago |
|
Caleb James DeLisle
|
b77de86321
Don't do the loop-route check until *after* splicing the route :|
|
11 years ago |
|
Caleb James DeLisle
|
e97d341f28
Improved logging
|
11 years ago |
|
Caleb James DeLisle
|
bb236d923c
merged next branch, step 1
|
11 years ago |
|
Caleb James DeLisle
|
fa15457db8
leak in sctive search counter caused searches not to be run after the node had run a long time.
|
11 years ago |
|
Caleb James DeLisle
|
409d6522bb
Removing searches from linked list incorrectly.
|
11 years ago |
|
Caleb James DeLisle
|
589c0375e5
Exported a new function to show the currently active searches and paid down some ugly technical debt
|
11 years ago |
|
Caleb James DeLisle
|
8634c1f75a
Off by one
|
11 years ago |
|
Caleb James DeLisle
|
17e62e7eeb
Fix bug causing occasional OOM with nodes using IPTunnel.
|
11 years ago |
 hartzler
|
3e6326c2fa
fixes signature of getClosest and clarifies test
|
11 years ago |
|
Caleb James DeLisle
|
e0541658ee
Start searches asynchronously to avoid returning null
|
11 years ago |
|
Caleb James DeLisle
|
6e9a7b7109
new approach to dealing with trashroutes
|
11 years ago |
