aboutsummaryrefslogtreecommitdiffstats
path: root/src/swarm/sync.rs
Commit message (Collapse)AuthorAgeFilesLines
* benchmark where twice comes fromtv3 days1-1/+1
|
* resend only what is neededtv3 days1-1/+1
|
* benchmark delivered and twicetv3 days1-0/+8
|
* any message taken as deliveredtv3 days1-11/+20
|
* tree messages need no answertv3 days1-2/+15
|
* writers from authorstv5 days1-6/+20
|
* tree changes had alreadytv5 days1-4/+14
|
* replace eager links by treestv5 days1-13/+5
|
* add writer treestv5 days1-6/+15
|
* introduce eager linkstv5 days1-5/+13
|
* document: introduce patch logtv6 days1-3/+3
|
* large messages over slow links are deltastv6 days1-7/+17
|
* peer worker sleeps until deadline, not fixed timetv6 days1-0/+11
|
* bootstrap: replace pull by progress notifytv6 days1-0/+1
|
* use Automerge as Document instead of AutoCommittv7 days1-5/+5
| | | | This allows a lot of operations to drop mut.
* move SyncSession to documenttv7 days1-1/+1
|
* move all automerge references to documenttv7 days1-14/+6
|
* move changes as bytes to documenttv7 days1-4/+1
|
* sync messages are bytes outside the sessiontv7 days1-3/+2
| | | | | | | | SyncPayload is now an opaque byte container in transport. SyncSession::generate returns the encoded payload, state::decode_sync(&payload) decodes it off the document lock (as before), and receive takes the decoded message, which is a type only the session can open. transport/mod.rs no longer imports Automerge.
* propagate clock figurestv8 days1-1/+4
|
* unify usage of is_dormanttv8 days1-2/+2
|
* unify request_sync logictv8 days1-21/+5
|
* move change publisher to own moduletv8 days1-81/+0
|
* move ChangeSet from swarm to documenttv8 days1-10/+1
|
* cleanup log messagestv11 days1-8/+8
|
* talk about "transport" unless mycelium is meanttv12 days1-3/+7
|
* reject and warn about too large delta blobstv12 days1-0/+5
|
* add delta blobstv12 days1-5/+116
|
* decode sync messages lazilytv12 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.
* add pull queuetv12 days1-2/+4
|
* send sync messages off workertv12 days1-1/+23
|
* add session IDstv13 days1-15/+23
| | | | | This prevents the regenerate flag from being lost across reset_sync during an in-flight send.
* wrap unix timestamps in newtypetv13 days1-1/+1
|
* SyncSessions owns Automerge's sync statetv13 days1-78/+11
|
* unify lock logictv13 days1-2/+2
|
* add clock moduletv13 days1-1/+2
| | | | | The Clock module is the system clock that every unix timestamp new goes through. It comes with an offset that only tests can set.
* recompute neighbours whenever topology changestv13 days1-0/+7
|
* late messages don't flip peer's epoch backtv13 days1-3/+4
|
* cache headstv2026-10-241-2/+3
| | | | To reduce the number of write locks, heads are now cached.
* being caught up needs some headstv2026-10-241-5/+9
| | | | | | | | | | | A sync message from a ready peer only completes the initial sync if that peer has at least one head. A peer with an empty document can no longer make a joiner ready on an empty store: a misconfigured second swarm, or a genesis node in its first seconds. Effect on a brand-new swarm: a joiner now waits until the genesis node has registered itself in the member list, which happens within about 5 s of it becoming ready.
* add chunked download and use it during rejointv2026-10-241-0/+6
|
* add throughput statstv2026-10-241-1/+18
|
* add propagation statstv2026-10-241-0/+15
|
* add swarm statstv2026-10-241-1/+7
|
* pull missing changestv2026-10-241-5/+14
| | | | | | Nodes push changes to their neighbours but don't relay them directly. In a partial mesh, nodes can learn about missing changes via keepalive messages and can pull them from their neighbours.
* sync with neighbourstv2026-10-241-1/+13
|
* retry using fresh messagestv2026-10-241-158/+127
| | | | | Don't store encoded messages but retry using a newly generated one. That way any accumulated changes can be delivered at once.
* drop stall breaker; resend needed changestv2026-10-241-65/+21
| | | | | | The stall breaker was superseded by regular re-handshaking Re-handshaking also gets triggered, when a node needs some of the heads its peer believes it has already sent. In that case the peer now resends what's needed.
* both sides may open sync after resettv2026-10-231-19/+8
| | | | | | | | | | Ready nodes will leave the first message to the pear that isn't ready. Hello and hello reply now carry ready, so readiness is known from the very first contact, before any sync. Together with the sync messages and keepalives, reports_ready is always current. After epoch switch, pings go to everyone except ready members already on the new epoch, these learn from the lagging peer's opening.
* perform handshake in one round triptv2026-10-231-1/+0
|