aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
...
* changes from a peer never go backtv5 days2-0/+52
|
* older delivered messages leave us leveltv5 days2-36/+76
|
* benchmark where twice comes fromtv5 days3-6/+20
|
* resend only what is neededtv5 days2-27/+62
|
* benchmark delivered and twicetv5 days3-0/+58
|
* delivered means has all up to our headstv5 days1-13/+59
|
* any message taken as deliveredtv5 days5-40/+112
|
* slow round back offtv5 days5-5/+92
|
* tree messages need no answertv5 days5-7/+235
|
* hello again when heard fromtv5 days3-2/+71
| | | | | | While we wait for a peer's HELLO_REPLY, anything else that comes from it makes our HELLO go again at once, at most once a second, instead of when the backoff says.
* bump pulled_changes_go_along_their_writers_treestv6 days1-0/+3
|
* unanswered pull frees the slottv6 days2-1/+23
| | | | | | When the answer to a pull is overdue (the sync reply timeout: 1 s in the simtests, 10 s in production), the pull gives up the slot, and the next waiting neighbour is pulled at once.
* no heal while both move ontv6 days1-3/+49
| | | | | | | The heal clock now also restarts when our heads and the peer's have both changed since the last STATUS that showed them differing. While writes go on, two neighbours's heads differ because of changes still on their way, and that alone no longer leads to a heal.
* bump pulled_changes_go_along_their_writers_treestv6 days1-2/+52
|
* overdue sync asked after from thentv6 days3-49/+92
|
* pull after gracetv6 days2-4/+33
| | | | | | | | Before pulling, a node waits one measured round trip to the announcer (1 s at most) and pulls only what is still missing. If the round trip isn't known yet, it pulls at once, as now. On the clean benchmark links the round trip is a few milliseconds, so those scenarios shouldn't change.
* announcements answered when lackingtv6 days5-12/+36
| | | | | | | A neighbour now answers an announcement only if it lacks some of the announced heads. Heads that merely differ are no longer enough, since they do while writes go on. The check reads the document, so it runs off the message consumer, as the pull check already does.
* unanswered sync asked aftertv6 days1-6/+28
| | | | | An overdue sync answer is now followed by a PING too, and the peer stays reached once it answers the PING.
* simtest: benchmark incidentstv6 days1-6/+21
|
* unanswered members not reachedtv6 days5-18/+132
|
* add member away benchmarktv6 days1-0/+34
|
* dashboard: show reach settv6 days1-2/+23
|
* links: only news pass ontv7 days1-6/+12
|
* dashboard: show writer treestv7 days3-26/+189
|
* add tree teststv7 days2-2/+144
|
* fixup 700, 701, 702tv7 days4-42/+43
|
* writers from authorstv7 days4-27/+68
|
* reach set for the sessiontv7 days4-19/+13
|
* add late arrivals benchmarktv7 days1-3/+49
|
* tree changes had alreadytv7 days2-4/+21
|
* announcement round counterstv7 days5-9/+46
|
* docs after treestv7 days2-9/+11
|
* announce what trees misstv7 days4-35/+119
|
* parallel propagation benchmarktv7 days1-14/+54
|
* quicker propagation benchmarktv7 days1-14/+21
|
* order members by distancetv7 days3-37/+63
|
* replace eager links by treestv7 days23-307/+48
|
* add writer treestv7 days11-33/+323
|
* add writer tree benchmarktv7 days1-0/+148
|
* measure 2-hop neighbourhoodtv7 days21-10/+230
|
* add propagation benchmarktv7 days3-1/+396
|
* dashboard: show eager links on the maptv7 days6-7/+65
|
* add eager link counterstv7 days4-0/+47
|
* introduce eager linkstv7 days14-24/+195
|
* skip neighbours with our headstv7 days3-15/+42
|
* announcements answered only if heads differtv7 days12-9/+46
|
* retry new peers before probingtv7 days2-6/+38
|
* probe list memberstv7 days4-9/+25
|
* links: add failovertv7 days1-14/+129
|
* swarm: links by whom we reachtv7 days3-12/+73
|