aboutsummaryrefslogtreecommitdiffstats
path: root/src/state
Commit message (Collapse)AuthorAgeFilesLines
* state: a document swap keeps the signatures of the changes it keepstv3 days1-3/+4
| | | | | | | | | | | | | Review 2.2: swap_document forgot every signature. After dropping changes taken unjudged (or a rejoin into the same epoch), the node could vouch for none of the changes it kept: under signatures required, those building on an actor waited for that actor's next signature, and for an actor that stopped writing, forever. Signatures::kept_for keeps the signatures of the changes the new document holds, checks each actor's latest again, and keeps those waiting for their changes. This node's replayed changes are signed at once, as before.
* sync: a member refused as a whole, synced with again once it changed documenttv3 days1-0/+5
| | | | | | | | | | | | A node refusing a member as a whole sent it nothing, and lifted that only when a change of the member's was taken again. But a member that dropped what was refused (REFUSALS) starts its sync sessions over, and a fresh session sends no changes until the other side answers: the two waited for each other, and the member stayed refused. The refusal now notes the member's run and epoch_seq when it began; a member with another document since (it dropped what was refused, or rejoined) is synced with again.
* access: a node judges its own changes before they leavetv3 days1-3/+51
| | | | | | | | | | | | | | | | | | | | The API checks the path a request asks for; the rules judge everything the change does. Where they disagreed, the honest node's change was refused by every other node, and with signatures required the node was refused as a whole. The plan had both gates run the same judgement. Under a policy, each change a node makes (its own writes and those through the API) is now judged by rules::judge as receivers will judge it, before it is signed or sent; one every node would refuse is rolled back (the document forked at the heads before it, same actor: the change never left) and the request refused with 403 (StoreError:: Refused). rules::judge_held, judging a change the document holds, moves from swarm::refusals to access::rules for both. The simulations that play a member past its own checks write with a helper that bypasses them (forge). Review finding 3 (the same judgement at both gates).
* access: an epoch's root is never droppedtv3 days1-2/+9
| | | | | | | | | | | | | | | A REFUSALS naming the swarm's first change made the node receiving it judge that change again in a document that isn't empty, refuse it as "a change without dependencies that isn't the epoch's root", and drop it with everything built on it: its whole document. The first epoch's root is now the first change a document holds, as it comes to an empty one or as it is in one; a change judged again on a peer's word is never refused for having no dependencies; and drop_taken refuses to drop an epoch's root, or anything that would leave nothing kept, logging an error instead. Review finding 1 (critical).
* swarm: changes taken unjudged, dropped (REFUSALS)tv4 days1-1/+85
| | | | | | | | | A node refusing changes a peer brought tells it which broke a rule; the peer judges them itself, and if its rules refuse them too, drops them with all built on them (its document forked without them, a fresh actor), quarantining its own writes among them. Wire version 42. Auth plan M6 (rejoin on taint).
* auth: logging in by keytv4 days1-0/+2
| | | | | | | | POST /auth/challenge and /auth/login/key: a node-signed challenge, valid five minutes and taken once, signed with one of the user's SSH keys (ssh-keygen -Y sign -n continuum-login), for a token. Auth plan M7 (key login, M2b).
* access: a node registers before it writestv4 days1-19/+83
| | | | | | | | | Under a policy, a change by a node that isn't a member is refused with all built on it: a node's registration is written before any write of its own or through the API, not after (the registration itself exempt). Auth plan M6 (fix).
* access: signatures requiredtv4 days2-3/+19
| | | | | | | | "signatures": "required": a change counts only once a signature by its committer vouches for it, and waits until then; every node then signs every change it makes. Auth plan M6.
* access: the policy enforcedtv4 days2-3/+29
| | | | | | | | _continuum/access/policy ({"version": 1}) has every node screen every change it receives, on every path changes come in. A policy a node doesn't know stops its writes through the API. GET /admin/refused. Auth plan M6.
* remake: forgetting and releasing name their usertv4 days1-0/+19
| | | | | | | Forget records and released quarantines name the user who made them, so they are judged by its capabilities. Auth plan M6.
* signatures: with sync messagestv4 days2-3/+12
| | | | | | | Sync messages carry, for each actor's writes, the latest signature the sender checked; the receiver checks it once it holds the change. Auth plan M5.
* signatures: --sign-everytv4 days1-3/+30
| | | | | | | | Changes left unsigned are vouched for by the node's next signature: --sign-every 1s (default) signs the latest within a second, all signs every change, none only those that must be. Auth plan M5.
* signatures: changes signed by their nodetv4 days2-5/+27
| | | | | | | | | | Every node signs the changes it commits with its node key, an Ed25519 signature over the change's hash, kept beside the document. One signature on an actor's latest change vouches for its earlier ones. Protocol data and paths listed in _continuum/access/signed are signed, and writes with ?signed=true. Auth plan M5.
* dashboard: without LINKS; LINKS removedtv4 days1-2/+0
| | | | | | | | | The dashboard reads every member's links from the members themselves; LINKS and LINKS_REPLY, swarm::links and their wire tags go (wire version bumped). Nodes the browser can't reach are drawn from what the reachable ones measure. Auth plan M3.
* dashboard: own links over the API, and CORStv4 days1-0/+2
| | | | | | | | | GET /admin/dashboard/links/own and its event stream: a node's own links, as LINKS_REPLY carried them, for other members' dashboards with a token. CORS for the members' dashboard origins and --cors-origins, without credentials. Auth plan M3.
* auth: logging in, and tokens as bearerstv4 days1-0/+4
| | | | | | | | | POST /auth/login (by htpasswd or the proxy), /auth/refresh, /auth/logout, GET /auth/whoami; every request may bring Authorization: Bearer instead of a password. --token-lifetime and --session-lifetime. Auth plan M2.
* access: first users at genesistv4 days1-0/+2
| | | | | | | | --bootstrap-users gives a new swarm a registry from its first change, each first user with everything; other users may do nothing until added. Without it the swarm has no registry and works as before. Auth plan M1.
* refuse misplaced authorstv5 days1-0/+5
|
* HyParView (slim)tv5 days1-4/+21
|
* dissemination traittv5 days1-0/+2
|
* topology traittv5 days1-0/+2
|
* handshakes countedtv5 days1-0/+56
|
* working links recordedtv5 days1-0/+2
|
* worker sends off the passtv6 days1-1/+6
|
* suspend detectedtv6 days1-0/+12
|
* one grace wait per peertv6 days1-0/+4
|
* asks by askertv6 days1-0/+5
|
* round trips in treestv6 days1-0/+5
|
* push by defaulttv6 days1-4/+0
|
* push changes experimenttv7 days1-0/+4
|
* announce only to those behindtv7 days1-2/+2
|
* asked again countertv7 days1-0/+5
|
* hop delaystv7 days1-0/+3
|
* slow round back offtv7 days2-0/+5
|
* hello again when heard fromtv7 days1-0/+52
| | | | | | 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.
* no heal while both move ontv7 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.
* overdue sync asked after from thentv7 days1-45/+88
|
* unanswered sync asked aftertv7 days1-6/+28
| | | | | An overdue sync answer is now followed by a PING too, and the peer stays reached once it answers the PING.
* unanswered members not reachedtv8 days2-3/+88
|
* announce what trees misstv8 days1-0/+3
|
* replace eager links by treestv8 days1-10/+0
|
* add writer treestv8 days2-3/+8
|
* measure 2-hop neighbourhoodtv8 days2-0/+7
|
* introduce eager linkstv9 days1-0/+10
|
* skip neighbours with our headstv9 days1-0/+5
|
* retry new peers before probingtv9 days1-5/+37
|
* probe list memberstv9 days1-3/+8
|
* add stream heartbeattv9 days1-0/+3
|
* dashboard: double-click opens peer dashboardtv9 days1-0/+3
|
* api: add x-continuum-heads to writestv10 days1-6/+10
|