aboutsummaryrefslogtreecommitdiffstats
path: root/src/swarm
Commit message (Collapse)AuthorAgeFilesLines
...
* get swarm RTT from pathstv9 days2-133/+91
|
* don't sample lost PONGstv9 days1-29/+23
|
* keepalive takes turnstv9 days3-3/+19
|
* move keepalive and firewall to pathstv9 days6-62/+196
|
* measure pathstv9 days1-1/+4
|
* rename PING/PONG to STATUS/STATUS_REPLYtv9 days9-70/+72
|
* build topology per transporttv10 days1-4/+128
|
* use libp2p mDNStv10 days1-0/+11
|
* add support for several transportstv10 days1-1/+7
|
* admit nodeId and addresses in the APItv10 days1-1/+8
|
* add identity layertv10 days1-2/+6
|
* introduce addressestv10 days4-16/+100
|
* introduce node keystv10 days1-6/+6
|
* talk about "transport" unless mycelium is meanttv10 days9-32/+37
|
* add smaller topology testtv10 days1-19/+35
|
* reject and warn about too large delta blobstv11 days2-0/+9
|
* add delta blobstv11 days3-19/+312
|
* make clippy happy againtv11 days5-35/+35
|
* warn about dropped blob requeststv11 days1-0/+28
|
* add sync and relay statstv11 days3-1/+44
|
* introduce bounded blob downloadstv11 days1-13/+41
|
* warn if swarm key is readable by otherstv11 days1-0/+22
|
* add farewell pingtv11 days2-2/+22
|
* decode sync messages lazilytv11 days1-1/+3
| | | | | | | Sync messages are now being decoded only where they are applied: after membership has been checked, in the sync lane's blocking task. Debug now shows only the size instead of the whole message.
* keepalive sends head samples alongside digesttv11 days1-21/+22
| | | | | | When there are more than 32 heads that can be send in a single keepalive message, the digest is accompanied by 32 heads, a different window each time, so after a few keepalives every head has been listed.
* add pull queuetv11 days3-34/+84
|
* send sync messages off workertv11 days1-1/+23
|
* read stamps sync and lock freetv11 days4-13/+20
|
* add sync lanestv11 days3-2/+70
|
* compute member pair scores cheaplytv11 days1-22/+65
|
* unreachability doesn't expiretv11 days2-9/+3
|
* forget records only oncetv11 days2-9/+37
|
* fix race between alive check and forgettingtv11 days2-8/+23
|
* forgotten nodes cancel their own recordstv11 days1-15/+14
|
* add session IDstv11 days1-15/+23
| | | | | This prevents the regenerate flag from being lost across reset_sync during an in-flight send.
* number handshakestv11 days2-7/+32
| | | | This prevents late hellos from resetting the session.
* keep silent strangers for a daytv11 days2-3/+40
|
* send bounded peer liststv11 days2-7/+38
|
* wrap unix timestamps in newtypetv11 days3-3/+3
|
* allow humand-friendly durations and sizes in CLItv11 days2-2/+2
|
* SyncSessions owns Automerge's sync statetv11 days2-80/+12
|
* unify keepalivetv11 days1-15/+23
|
* split peer moduletv11 days9-915/+995
|
* unify peers snapshot logictv11 days1-11/+3
|
* unify lock logictv11 days10-70/+31
|
* move protocol paths to dedicated moduletv11 days3-7/+7
|
* refresh peers after stalltv11 days2-1/+19
|
* add random swarms testtv11 days1-0/+62
|
* add clock moduletv11 days3-6/+11
| | | | | The Clock module is the system clock that every unix timestamp new goes through. It comes with an offset that only tests can set.
* avoid stalled sourcetv11 days1-2/+24
| | | | | | | | | | | | | A source given up for not answering our sync messages could become the source again straight away. Whoever asks first after the reset takes the role, and that's often the same peer, which would then stall for another delivery limit. Now, a stalled source isn't chosen again for one more delivery_limit. If it's the only candidate, catching up waits that long and then tries it again. Before, a retry with that peer couldn't have helped any sooner anyway: a source is only given up after a whole delivery limit without a sync message, and that clock restarts when it is chosen again.