aboutsummaryrefslogtreecommitdiffstats
path: root/src/swarm
Commit message (Collapse)AuthorAgeFilesLines
...
* push only below deflatetv13 days1-2/+8
|
* push changes experimenttv13 days1-16/+78
|
* suspect pathstv13 days1-0/+5
|
* last chunks from twotv13 days1-1/+55
|
* unreached sources hand overtv13 days1-1/+26
|
* chunk timeout adaptstv13 days1-6/+74
|
* manifests asked at oncetv13 days1-45/+57
|
* measure deflated chunkstv13 days2-8/+28
|
* deflate blob chunkstv13 days2-5/+29
|
* deflate blob chunkstv13 days2-5/+44
|
* epoch tag in synctv13 days2-9/+34
|
* no swarm ID in synctv13 days2-24/+9
|
* announce only to those behindtv13 days2-12/+31
|
* measure sync envelopestv13 days2-0/+24
|
* dashboard: swarm-wide epoch shinetv14 days1-8/+31
|
* asked known elsewheretv14 days3-2/+23
|
* predecessors go alongtv14 days1-3/+18
|
* covered neighbours have ittv14 days3-16/+54
|
* new changes per messagetv14 days2-1/+10
|
* neigbours by transporttv14 days1-1/+21
|
* own tree shapetv14 days1-10/+136
|
* drop full mesh stattv14 days1-3/+7
|
* delta from held headstv14 days1-1/+4
|
* offer again ungeneratedtv14 days1-32/+53
|
* stale delta offerstv14 days1-0/+19
|
* blob window growstv14 days1-5/+10
|
* delta known to sendertv14 days1-1/+7
|
* delta propagationtv14 days2-2/+18
|
* one delta at a timetv14 days2-9/+53
|
* time blob messagestv14 days4-4/+46
|
* log why a deltatv14 days1-0/+6
|
* offered deltas lapsetv14 days1-4/+14
|
* throughput from delivered messagetv14 days2-2/+59
|
* hop delaystv14 days4-2/+128
|
* benchmark delays by pathtv14 days2-0/+15
|
* fix: change publisher stops with the runtimetv14 days1-3/+8
|
* benchmark what sync bytes aretv14 days2-3/+30
|
* changes from a peer never go backtv14 days1-0/+4
|
* older delivered messages leave us leveltv14 days1-1/+3
|
* benchmark where twice comes fromtv14 days2-4/+11
|
* resend only what is neededtv14 days1-1/+1
|
* benchmark delivered and twicetv14 days2-0/+30
|
* any message taken as deliveredtv14 days2-13/+23
|
* slow round back offtv14 days2-5/+86
|
* tree messages need no answertv14 days3-6/+40
|
* hello again when heard fromtv14 days2-2/+19
| | | | | | 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.
* unanswered pull frees the slottv2026-11-042-1/+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.
* overdue sync asked after from thentv2026-11-042-4/+4
|
* pull after gracetv2026-11-042-4/+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 lackingtv2026-11-044-10/+34
| | | | | | | 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.