aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* remake coordinators resume after restarttv13 days3-5/+49
|
* don't commit expired remakestv13 days1-58/+93
|
* add votes registertv13 days8-124/+507
|
* bound stable-hub degreetv13 days1-4/+46
| | | | | | Each stable member can be preferred at most by 4*k members. The budget is spent in key order, which is the same on every node, so the graph stays deterministic, symmetric and connected.
* coalesce announcementstv13 days3-2/+60
|
* cache headstv13 days7-24/+61
| | | | To reduce the number of write locks, heads are now cached.
* corroberate newer epochstv13 days1-11/+53
|
* follow and download from ready members onlytv13 days3-3/+29
|
* being caught up needs some headstv13 days1-5/+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.
* only protocol-writers may modify _continuumtv13 days6-14/+63
|
* put remake content into blobstv13 days15-79/+319
|
* add chunked download and use it during rejointv13 days15-3/+844
|
* allow unsetting config optionstv13 days2-46/+264
|
* add API and CLI to print effective configtv13 days4-2/+106
|
* add configuration filestv13 days4-107/+1261
|
* wire up --mycelium-tcp-listen-porttv13 days1-2/+5
| | | | | It wasn't used before, causing the supervised daemon to always pick a random one.
* introduce majority voting on remakestv13 days7-95/+406
|
* add electoratetv13 days3-8/+109
| | | | | For an upcoming change in remake coordination, nodes now maintain their self-assessed liveness in the document.
* aliases are not in logs (fixup)tv13 days2-2/+2
|
* measure document saved-size only after changestv13 days2-18/+43
|
* allow import remake using the raw documenttv13 days1-6/+85
|
* add the import kind of remaketv13 days3-49/+200
|
* add endpoint to export the documenttv13 days3-0/+88
| | | | GET /admin/document to retrieve the automerge document.
* add node aliasestv13 days6-11/+147
|
* expose stats for prometheus to scrapetv13 days2-7/+514
|
* add throughput statstv13 days8-2/+231
|
* add propagation statstv13 days4-1/+126
|
* add per peer incident statstv13 days4-4/+52
|
* expose stats via the APItv13 days2-0/+136
| | | | GET /admin/stats to retrieve all stats.
* add swarm statstv13 days9-1/+238
|
* add traffic countertv14 days9-12/+270
| | | | | | | | | | This introduces a new wrapper that counts messages and payload bytes per peer, split by kind (hello, helloReply, sync, ping, pong), each sent, received and refused. The traffic counter goes around the firewall, which means messages dropped outbound are still counted as sent, and messages dropped inbound don't count as received.
* nodes prefer stable neighbourstv14 days1-10/+89
|
* assess swarm stabilitytv14 days8-11/+330
|
* record RTT sampling instanttv14 days4-30/+104
|
* prefer closer sources to catch uptv14 days3-3/+49
|
* record RTT from keepalive messagestv14 days7-20/+148
|
* pull missing changestv14 days7-5/+162
| | | | | | 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.
* find reachable neighbourstv14 days3-55/+162
|
* sync with neighbourstv14 days6-4/+171
|
* add neighbour graphtv14 days2-0/+180
|
* add member registrytv14 days4-1/+73
|
* use nonce in handshaketv14 days5-2/+39
| | | | | To prevent stale hello replies to cause joins and thus stale sessions, each run gets a random ID, which used as nonce in the handshake.
* drain queued messages on startuptv14 days1-2/+48
|
* consolidate backoff codetv14 days4-49/+99
|
* retry using fresh messagestv14 days3-170/+180
| | | | | Don't store encoded messages but retry using a newly generated one. That way any accumulated changes can be delivered at once.
* heal from noticed divergencetv14 days1-6/+27
| | | | | Wait 2*keepalive but at least 20 s to trigger resync so that lost messages can be resent first.
* heal after reply timeouttv14 days1-0/+3
|
* drop stall breaker; resend needed changestv14 days2-82/+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.
* no heal during catch-uptv14 days1-1/+6
|
* peers that just became ready get syncstv14 days1-0/+9
| | | | | | When keepalive says ready and we didn't have taht on record before, the peer is queued for sync. Hello and hello reply don't do this because the handshake already queues a sync.