refactor(#2808): the wire state enum is the scheduler's own

`hive_host_sock::jobs::State` was a hand-maintained copy of
`hive_jobq::State` — five variants spelled the same in both, kept in sync
by whoever remembered. Adding `Skipped` last week meant adding it twice.
The wire crate now re-exports the scheduler's enum and `to_wire_state` is
gone.

Two states that were hidden now reach clients. `to_wire_state` renamed
`Pending` to `Queued` and folded `Finishing` into `Running`, so the
dashboard could not distinguish a node waiting on its dependencies from
one whose own work is done while its sub-nodes still run. Both are now
visible, and consumers say which they mean.

Every consumer had to move with it, and only the Rust ones said so: the
exhaustive matches in `hivectl` and `DagView::rollup_state` failed to
compile, while the dashboard's fourteen string comparisons would have
gone quietly wrong — a `finishing` node no longer counting as running,
a `pending` node no longer as queued.

The frontend also builds CSS class names out of the state string
(`rqe-` + state, `rqe-node-` + state) and keys its glyph map on it, all
lowercase. Those go through a `stateSlug` helper now; comparisons use the
wire spelling, presentation lowercases. Without that split every queue
entry and node chip would have silently lost its styling.

Dropping the `State as JobState` alias in hive-c0re falls out of this:
the alias only existed to tell two `State` types apart, and there is one
now.
This commit is contained in:
atlas 2026-07-27 21:50:24 +02:00
commit b9aab7e923
8 changed files with 72 additions and 93 deletions

1
Cargo.lock generated
View file

@ -1716,6 +1716,7 @@ name = "hive-host-sock"
version = "0.1.0"
dependencies = [
"chrono",
"hive-jobq",
"hive-sh4re",
"hive-types",
"serde",

View file

@ -25,20 +25,28 @@ let rebuildQueueState = [];
// DagView no longer carries top-level `kind`/`state`/`started_at`/`finished_at`
// — these are all derived from the NodeView array by the client.
// Rollup state from nodes (failed > cancelled > running > queued).
// Node/DAG states arrive in the wire spelling of `hive_jobq::State` — the
// scheduler's own enum, serialised verbatim, so the names are PascalCase and
// there is no separate display-shaped wire type. Compare against those names;
// lowercase only where a CSS class or human-facing label needs it.
const stateSlug = (s) => String(s || '').toLowerCase();
// Rollup state from nodes (Failed > Cancelled > Running > Pending).
// `Done` nodes are excluded from the payload, so a fully-done DAG is absent;
// an empty nodes array should not arise in practice — return 'done' defensively.
// an empty nodes array should not arise in practice — return 'Done' defensively.
function rollupState(nodes) {
const ns = nodes || [];
if (!ns.length) return 'done';
if (ns.some((n) => n.state === 'failed')) return 'failed';
if (ns.some((n) => n.state === 'cancelled')) return 'cancelled';
if (ns.some((n) => n.state === 'running')) return 'running';
if (!ns.length) return 'Done';
if (ns.some((n) => n.state === 'Failed')) return 'Failed';
if (ns.some((n) => n.state === 'Cancelled')) return 'Cancelled';
// `Finishing` is a node whose own work is done while its sub-nodes still
// run — in flight, so it counts as running.
if (ns.some((n) => n.state === 'Running' || n.state === 'Finishing')) return 'Running';
// A skipped node is a branch the run ruled out, which is expected on a
// healthy DAG — it must not make the roll-up read as still-pending. This
// mirrors `DagView::rollup_state` in hive-host-sock; edit the two together.
if (ns.every((n) => n.state === 'skipped' || n.state === 'done')) return 'done';
return 'queued';
if (ns.every((n) => n.state === 'Skipped' || n.state === 'Done')) return 'Done';
return 'Pending';
}
@ -168,7 +176,7 @@ const QUEUE_STATE_GLYPH = {
};
function firstFailedNode(entry) {
return (entry.nodes || []).find((n) => n.state === 'failed') || null;
return (entry.nodes || []).find((n) => n.state === 'Failed') || null;
}
// Topo-sort a flat node list using `deps` edges. Nodes whose deps are all
@ -299,22 +307,23 @@ function renderQueueEntry(entry) {
const finishedAt = isoToSecs(entry.finished_at);
const createdAt = isoToSecs(entry.created_at);
const slug = stateSlug(state);
const li = el('li', {
class: 'rebuild-queue-entry rqe-' + state,
class: 'rebuild-queue-entry rqe-' + slug,
'data-id': String(entry.id),
});
li.append(
el('span', { class: 'rqe-state', title: state }, QUEUE_STATE_GLYPH[state] || '?'),
el('span', { class: 'rqe-state', title: slug }, QUEUE_STATE_GLYPH[slug] || '?'),
' ',
el('span', { class: 'rqe-kind' }, entry.source),
);
li.append(' ', el('span', { class: 'rqe-source rqe-source-' + entry.source }, entry.source));
if (state === 'queued') {
if (state === 'Pending') {
li.append(' ', el('span', {
class: 'rqe-when',
'data-rqe-enqueued': String(createdAt ?? ''),
}, '· queued ' + (createdAt ? fmtAgo(createdAt) : '')));
} else if (state === 'running' && startedAt) {
} else if (state === 'Running' && startedAt) {
const elapsed = Math.max(0, Math.floor(Date.now() / 1000) - startedAt);
li.append(' ', el('span', {
class: 'rqe-when',
@ -324,8 +333,8 @@ function renderQueueEntry(entry) {
li.append(' ', el('span', {
class: 'rqe-when',
'data-rqe-finished': String(finishedAt),
'data-rqe-state': state,
}, '· ' + state + ' ' + fmtAgo(finishedAt)));
'data-rqe-state': slug,
}, '· ' + slug + ' ' + fmtAgo(finishedAt)));
}
if (entry.reason) {
const r = entry.reason.split('\n')[0];
@ -359,7 +368,7 @@ function renderQueueEntry(entry) {
}));
}
const chip = el('span', {
class: 'rqe-node rqe-node-' + n.state,
class: 'rqe-node rqe-node-' + stateSlug(n.state),
title: (n.agent ? n.agent + ' · ' : '') + n.kind + ' · ' + n.state
+ (n.error ? ' — ' + n.error : ''),
}, (QUEUE_STATE_GLYPH[n.state] || '?') + ' ' + n.kind);
@ -390,7 +399,7 @@ function renderQueueEntry(entry) {
if (failed && failed.error) {
li.append(el('pre', { class: 'rqe-error', title: failed.error }, truncate(failed.error, 200)));
}
if (state === 'queued') {
if (state === 'Pending') {
const cancelForm = el('form', {
method: 'POST',
action: '/api/rebuild-queue/' + entry.id + '/cancel',
@ -421,7 +430,7 @@ function renderQueueEntry(entry) {
// rebuild_queue_changed tick; a terminal node's log is static (one last fetch
// on transition, then done).
let liveLogId = null; // current node id being shown
let liveLogDone = false; // true once the node left 'running'
let liveLogDone = false; // true once the node left 'Running'
let liveLogCollapsed = false;
let liveLogPollTimer = null;
@ -433,8 +442,8 @@ function clearLiveLogPoll() {
// Gate on has_log so lock/noop/store-only nodes don't open a blank panel.
function findLiveBuild(queue) {
for (const e of queue || []) {
if (rollupState(e.nodes || []) !== 'running') continue;
const node = (e.nodes || []).find((n) => n.state === 'running' && n.has_log);
if (rollupState(e.nodes || []) !== 'Running') continue;
const node = (e.nodes || []).find((n) => n.state === 'Running' && n.has_log);
if (node) return { entry: e, node };
}
return null;
@ -471,14 +480,14 @@ function renderRebuildLiveLog(queue) {
// Same node, already polling — just let the timer tick (or do a final
// fetch if the node just went non-running and we haven't marked done yet).
if (liveNode.id === liveLogId) {
if (!liveLogDone && liveNode.state !== 'running') {
if (!liveLogDone && liveNode.state !== 'Running') {
clearLiveLogPoll();
liveLogDone = true;
const pre = root.querySelector('.rebuild-live-log-output');
const badge = root.querySelector('.rebuild-live-log-badge');
if (pre) fetchAndRenderLiveLog(liveNode.id, pre);
if (badge) {
const ok = liveNode.state !== 'failed';
const ok = liveNode.state !== 'Failed';
badge.className = 'rebuild-live-log-badge ' + (ok ? 'rll-ok' : 'rll-fail');
badge.textContent = liveNode.state;
}
@ -536,7 +545,7 @@ function updateRebuildCount() {
let n = 0;
for (const e of rebuildQueueState) {
const s = rollupState(e.nodes || []);
if (s === 'queued' || s === 'running') n++;
if (s === 'Pending' || s === 'Running') n++;
}
if (n > 0) { pill.textContent = String(n); pill.hidden = false; }
else { pill.hidden = true; }

View file

@ -75,7 +75,7 @@ export function applyRebuildQueueChanged(ev) {
function inFlightOpsByAgent() {
const out = new Map();
for (const e of rebuildQueueState) {
if (e.state !== 'queued' && e.state !== 'running') continue;
if (e.state !== 'Pending' && e.state !== 'Running') continue;
// spawn ops target an agent that doesn't exist yet as a
// container — the transient store already drives the
// pending row for that case. Skip here to avoid double-
@ -85,15 +85,15 @@ function inFlightOpsByAgent() {
const perAgentState = new Map();
for (const n of e.nodes || []) {
if (!n.agent) continue;
if (n.state !== 'queued' && n.state !== 'running') continue;
if (n.state !== 'Pending' && n.state !== 'Running') continue;
const cur = perAgentState.get(n.agent);
if (!cur || (cur === 'queued' && n.state === 'running')) {
if (!cur || (cur === 'Pending' && n.state === 'Running')) {
perAgentState.set(n.agent, n.state);
}
}
for (const [agent, state] of perAgentState) {
const cur = out.get(agent);
if (!cur || (cur.state === 'queued' && state === 'running')) {
if (!cur || (cur.state === 'Pending' && state === 'Running')) {
out.set(agent, { kind: e.kind, state });
}
}
@ -774,10 +774,10 @@ export function renderContainers(s) {
// running step is already visible per-agent on each card (transient +
// in-flight-queue badges), so the top of the tab only needs the summary.
const activeQueue = rebuildQueueState.filter(
(e) => e.state === 'queued' || e.state === 'running',
(e) => e.state === 'Pending' || e.state === 'Running',
);
if (activeQueue.length) {
const running = activeQueue.filter((e) => e.state === 'running').length;
const running = activeQueue.filter((e) => e.state === 'Running').length;
const queued = activeQueue.length - running;
const parts = [];
if (running) parts.push(`${running} running`);
@ -840,7 +840,7 @@ export function renderContainers(s) {
const transientKind = transientsState.get(c.name)?.kind || null;
const op = !transientKind ? inFlight.get(c.name) : null;
const pending = transientKind
|| (op && (op.state === 'running'
|| (op && (op.state === 'Running'
? (op.kind === 'meta_update' ? 'meta-updating'
: op.kind === 'destroy' ? 'destroying'
: op.kind === 'restart' ? 'restarting'
@ -858,7 +858,7 @@ export function renderContainers(s) {
: op.kind === 'reconcile' ? 'reconcile queued'
: 'rebuild queued')));
const opRunning = transientKind != null
|| (op != null && op.state === 'running');
|| (op != null && op.state === 'Running');
const selected = selectionState.has(c.name);
// Pending questions where this agent is the asker (awaiting an
// answer) or the target (owes a reply). Derived live from

View file

@ -43,7 +43,7 @@ use chrono::{DateTime, Utc};
use hive_host_sock::jobs::NodeView;
use hive_jobq::resources::ResourceTable;
use hive_jobq::scheduler::{Outcome, Scheduler};
use hive_jobq::{Dep, Graph, NodeId, State as JobState};
use hive_jobq::{Dep, Graph, NodeId};
use hive_sh4re::wire_time::now_unix;
use tokio::sync::Notify;
@ -133,19 +133,6 @@ impl Default for JobQueue {
}
}
/// Map a crate node state onto the wire state (`Pending` ↔ `Queued`;
/// `Finishing` — own logic done, sub-nodes still running — reads as `Running`).
fn to_wire_state(state: JobState) -> State {
match state {
JobState::Pending => State::Queued,
JobState::Running | JobState::Finishing => State::Running,
JobState::Done => State::Done,
JobState::Failed => State::Failed,
JobState::Cancelled => State::Cancelled,
JobState::Skipped => State::Skipped,
}
}
/// Insert `nodes` into the shared graph, honouring the spec's explicit **parent
/// axis**: a node with `parent = None` is a top-level group root (re-parented to
/// `group_parent`, which is `None` for `submit` and the emitting node for
@ -476,7 +463,7 @@ impl QueueInner {
self.sched
.graph()
.node(id)
.is_some_and(|n| n.state == JobState::Running)
.is_some_and(|n| n.state == State::Running)
}
/// The container node of `dag_id` — the `NodeKind::Dag` root whose id equals
@ -548,10 +535,10 @@ impl QueueInner {
// `Done` nodes drop off the wire — a finished step isn't
// interesting. `Skipped` ones stay: which branch a run *didn't*
// take is the readable half of an outcome-branched DAG.
if matches!(node.state, JobState::Done) {
if matches!(node.state, State::Done) {
continue;
}
any_unsettled |= !matches!(node.state, JobState::Skipped);
any_unsettled |= !matches!(node.state, State::Skipped);
let deps: Vec<u64> = node
.deps
.iter()
@ -587,7 +574,7 @@ impl QueueInner {
agent: node.payload.agent().to_owned(),
kind: node.payload.as_str().to_owned(),
deps,
state: to_wire_state(node.state),
state: node.state,
started_at: node.started_at,
finished_at: node.finished_at,
error: node.error.clone(),

View file

@ -9,6 +9,7 @@ workspace = true
[dependencies]
chrono.workspace = true
hive-jobq.workspace = true
hive-sh4re.workspace = true
hive-types.workspace = true
serde.workspace = true

View file

@ -40,36 +40,7 @@ impl Source {
}
}
/// Lifecycle state of a node — and, rolled up, of a DAG.
#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "snake_case")]
pub enum State {
Queued,
Running,
Done,
Failed,
Cancelled,
/// The node's own dependency edges ruled it out: an outcome branch that
/// wasn't taken. Distinct from `Cancelled`, which is work actively
/// dropped. A skipped node is an expected part of a healthy run — every
/// approval DAG has two not-taken tails and every rebuild has one — so
/// consumers must not read it as a failure or cancellation signal.
Skipped,
}
impl State {
/// Whether the node will never change state again. `Skipped` counts:
/// a branch that was ruled out is as final as one that ran, and the
/// wait loops (`hivectl`'s progress display, the daemon's
/// dag-settled check) hang forever if it doesn't.
#[must_use]
pub fn is_terminal(self) -> bool {
matches!(
self,
State::Done | State::Failed | State::Cancelled | State::Skipped
)
}
}
pub use hive_jobq::State;
/// Kind-specific payload for `Template::PermChange` DAGs.
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
@ -187,17 +158,20 @@ impl DagView {
/// Rust consumer (hivectl, the wait loops, tests) uses so the dashboard's
/// JS render and the host agree: `Failed` if any node failed, else
/// **`Cancelled` if any cancelled**, else `Running` if any running, else
/// `Queued` if any queued, else `Done`. `Done` nodes are excluded
/// `Pending` if any pending, else `Done`. `Done` nodes are excluded
/// from the wire, so a DAG that is *entirely* done isn't sent at all —
/// its absence from the snapshot is what signals completion.
///
/// `Cancelled` outranks both `Running` and `Queued` because a cancelled DAG
/// `Cancelled` outranks both `Running` and `Pending` because a cancelled DAG
/// still has its weak-edged tail node to run (it reports the cancellation),
/// so `Queued`-then-`Running` would flicker back at the operator who just
/// so `Pending`-then-`Running` would flicker back at the operator who just
/// cancelled it and read as "the cancel didn't take". Outside that window
/// the states barely co-occur: a cancel *cascade* originates at a `Failed`
/// node, which returns early above.
///
/// `Finishing` counts as running: the node's own work is done but its
/// sub-nodes are still going, so the DAG is still in flight.
///
/// `Skipped` contributes nothing: a not-taken branch is an expected part of
/// a healthy run, so counting it would make every successful DAG roll up
/// non-`Done`.
@ -208,13 +182,13 @@ impl DagView {
#[must_use]
pub fn rollup_state(&self) -> State {
let mut any_running = false;
let mut any_queued = false;
let mut any_pending = false;
let mut any_cancelled = false;
for n in &self.nodes {
match n.state {
State::Failed => return State::Failed,
State::Running => any_running = true,
State::Queued => any_queued = true,
State::Running | State::Finishing => any_running = true,
State::Pending => any_pending = true,
State::Cancelled => any_cancelled = true,
State::Done | State::Skipped => {}
}
@ -223,8 +197,8 @@ impl DagView {
State::Cancelled
} else if any_running {
State::Running
} else if any_queued {
State::Queued
} else if any_pending {
State::Pending
} else {
State::Done
}

View file

@ -205,6 +205,11 @@ pub enum Dep<R> {
}
/// A node's lifecycle state.
///
/// This type *is* the wire representation — `hive-host-sock` hands it to
/// clients verbatim rather than mapping it through a parallel enum — so the
/// serialised names (`"Pending"`, `"Running"`, …) are what every consumer,
/// including the dashboard's JS, matches on.
#[derive(Debug, Clone, Copy, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub enum State {
/// Waiting on dependencies (node or resource).
@ -213,9 +218,9 @@ pub enum State {
Running,
/// Own logic finished successfully, but the node is *not yet terminal*: it
/// waits here until all its sub-nodes ([`Node::parent`] children) are
/// terminal, then rolls up to [`State::Done`] (every child `Done`) or
/// [`State::Failed`] (any child `Failed`/`Cancelled`). A node with no
/// children never rests here — it goes straight to a terminal state.
/// terminal, then rolls up — `Failed` if any child failed, else `Cancelled`
/// if any was dropped, else `Done`. A node with no children never rests
/// here — it goes straight to a terminal state.
Finishing,
/// Completed successfully — own logic done *and* every sub-node `Done`.
Done,

View file

@ -286,8 +286,10 @@ fn node_line(
fn state_glyph(state: hive_host_sock::jobs::State) -> &'static str {
match state {
hive_host_sock::jobs::State::Queued => "",
hive_host_sock::jobs::State::Running => "",
hive_host_sock::jobs::State::Pending => "",
// `Finishing` is own-work-done with sub-nodes still going — in flight,
// so it reads the same as running.
hive_host_sock::jobs::State::Running | hive_host_sock::jobs::State::Finishing => "",
hive_host_sock::jobs::State::Done => "",
hive_host_sock::jobs::State::Failed => "",
hive_host_sock::jobs::State::Cancelled => "",
@ -362,7 +364,7 @@ mod tests {
node(0, "alice", "prebuild", State::Done),
node(1, "alice", "stop_for_update", State::Done),
node(2, "alice", "swap", State::Running),
node(3, "alice", "reconcile", State::Queued),
node(3, "alice", "reconcile", State::Pending),
],
};
let line = render_dag_line(&dag);