aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* dashboard: hop delaystv9 days1-3/+35
|
* throughput from delivered messagetv9 days2-2/+59
|
* hop delaystv9 days8-3/+178
|
* test links fail overtv9 days1-0/+105
|
* benchmark members joiningtv9 days1-5/+48
|
* benchmark delays by pathtv9 days3-0/+28
|
* tests count the changes senttv9 days1-5/+7
|
* fix: change publisher stops with the runtimetv9 days1-3/+8
|
* no reset for heads still on their waytv9 days1-1/+44
|
* benchmark what sync bytes aretv9 days4-3/+74
|
* changes from a peer never go backtv9 days2-0/+52
|
* older delivered messages leave us leveltv9 days2-36/+76
|
* benchmark where twice comes fromtv9 days3-6/+20
|
* resend only what is neededtv9 days2-27/+62
|
* benchmark delivered and twicetv9 days3-0/+58
|
* delivered means has all up to our headstv9 days1-13/+59
|
* any message taken as deliveredtv9 days5-40/+112
|
* slow round back offtv9 days5-5/+92
|
* tree messages need no answertv9 days5-7/+235
|
* hello again when heard fromtv9 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_treestv10 days1-0/+3
|
* unanswered pull frees the slottv10 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 ontv10 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_treestv10 days1-2/+52
|
* overdue sync asked after from thentv10 days3-49/+92
|
* pull after gracetv10 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 lackingtv10 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 aftertv10 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 incidentstv10 days1-6/+21
|
* unanswered members not reachedtv11 days5-18/+132
|
* add member away benchmarktv11 days1-0/+34
|
* dashboard: show reach settv11 days1-2/+23
|
* links: only news pass ontv11 days1-6/+12
|
* dashboard: show writer treestv11 days3-26/+189
|
* add tree teststv11 days2-2/+144
|
* fixup 700, 701, 702tv11 days4-42/+43
|
* writers from authorstv11 days4-27/+68
|
* reach set for the sessiontv11 days4-19/+13
|
* add late arrivals benchmarktv11 days1-3/+49
|
* tree changes had alreadytv11 days2-4/+21
|
* announcement round counterstv11 days5-9/+46
|
* docs after treestv11 days2-9/+11
|
* announce what trees misstv11 days4-35/+119
|
* parallel propagation benchmarktv11 days1-14/+54
|
* quicker propagation benchmarktv11 days1-14/+21
|
* order members by distancetv11 days3-37/+63
|
* replace eager links by treestv11 days23-307/+48
|
* add writer treestv11 days11-33/+323
|
* add writer tree benchmarktv11 days1-0/+148
|
* measure 2-hop neighbourhoodtv11 days21-10/+230
|
generated by cgit v1.3.1 (git 2.54.0) at 2026-11-13 20:11:24 +0000