aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* benchmark delivered and twicetv6 days3-0/+58
|
* delivered means has all up to our headstv6 days1-13/+59
|
* any message taken as deliveredtv6 days6-47/+121
|
* slow round back offtv6 days7-10/+102
|
* tree messages need no answertv6 days6-10/+245
|
* hello again when heard fromtv6 days4-3/+75
| | | | | | 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_treestv7 days1-0/+3
|
* unanswered pull frees the slottv7 days3-4/+27
| | | | | | 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 ontv7 days3-5/+51
| | | | | | | 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_treestv7 days1-2/+52
|
* overdue sync asked after from thentv7 days3-49/+92
|
* pull after gracetv7 days3-5/+36
| | | | | | | | 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 lackingtv7 days7-15/+40
| | | | | | | 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 aftertv7 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 incidentstv7 days1-6/+21
|
* unanswered members not reachedtv8 days7-24/+143
|
* add member away benchmarktv8 days2-1/+36
|
* dashboard: show reach settv8 days2-2/+24
|
* links: only news pass ontv8 days1-6/+12
|
* dashboard: show writer treestv8 days5-26/+233
|
* add tree teststv8 days2-2/+144
|
* update architecture doctv8 days1-4/+7
|
* fixup 700, 701, 702tv8 days4-42/+43
|
* writers from authorstv8 days5-32/+75
|
* reach set for the sessiontv8 days6-22/+18
|
* add late arrivals benchmarktv8 days1-3/+49
|
* tree changes had alreadytv8 days3-5/+24
|
* announcement round counterstv8 days7-10/+53
|
* docs after treestv8 days3-10/+12
|
* announce what trees misstv8 days5-36/+125
|
* parallel propagation benchmarktv8 days2-15/+57
|
* quicker propagation benchmarktv8 days2-15/+23
|
* order members by distancetv8 days4-41/+69
|
* replace eager links by treestv8 days28-339/+61
|
* add writer treestv8 days12-33/+339
|
* add writer tree benchmarktv8 days2-1/+152
|
* measure 2-hop neighbourhoodtv8 days23-11/+250
|
* add propagation benchmarktv8 days4-1/+404
|
* dashboard: show eager links on the maptv8 days8-11/+74
|
* add eager link counterstv8 days6-1/+57
|
* introduce eager linkstv8 days16-25/+210
|
* skip neighbours with our headstv8 days5-17/+47
|
* announcements answered only if heads differtv8 days14-12/+50
|
* retry new peers before probingtv8 days3-7/+41
|
* probe list memberstv8 days5-10/+28
|
* links: add failovertv8 days2-14/+135
|
* swarm: links by whom we reachtv8 days4-13/+75
|
* always announce new headstv8 days5-52/+68
|
* dashboard: neater graph title & legendtv8 days2-25/+46
|
* dashboard: legend doesn't interfere with graphtv8 days2-5/+17
|