Judge each machine's report from its own send, so a first machine opens the gate

With one machine first a module is sent twice, and the tier gate asked every
machine for a report after the second send: the first machine's report, made
between the two, read as stale and the plan waited for ever (hq issue 256).
Also round a plan's wait to the second, not the minute, so it is not 0s.
This commit is contained in:
jochen
2026-10-05 18:29:59 +02:00
parent bdfbd0254f
commit 0d2b304f08
2 changed files with 65 additions and 4 deletions
+34 -4
View File
@@ -282,6 +282,22 @@ func gates(p inventory.Plan, edges []inventory.Edge, rollsOut func(string) bool)
}
// applied says whether every machine running the module has reported since the module was built.
// appliedEach is applied with a moment of its own for each machine: the reports that count are the ones
// after that machine was sent the build (novox/hq issue 256).
func appliedEach(module string, since func(node string) time.Time, running []string, reports []inventory.Reported) (bool, []string) {
at := map[string]*time.Time{}
for _, r := range reports {
at[r.Node] = r.At
}
var waiting []string
for _, n := range running {
if t := at[n]; t == nil || t.Before(since(n)) {
waiting = append(waiting, n)
}
}
return len(waiting) == 0, waiting
}
func applied(module string, builtAt time.Time, running []string, reports []inventory.Reported) (bool, []string) {
at := map[string]*time.Time{}
for _, r := range reports {
@@ -637,10 +653,24 @@ func advanceOnce(ctx context.Context, open *stores, p *inventory.Plan,
if state.BuiltAt != nil {
since = *state.BuiltAt
}
if state.SentAt != nil && state.SentAt.After(since) {
since = *state.SentAt
// **Each machine from its own send** (novox/hq issue 256). With one machine first (ADR 0218)
// a module is sent twice — the first machine, then the rest — and SentAt is the second.
// Asked of every machine, the first machine's report, made between the two sends, read as
// older than the build, and the gate waited for a report it already had, for ever.
sinceFor := func(node string) time.Time {
at := since
sent := state.SentAt
for _, n := range state.First {
if n == node {
sent = state.FirstAt
}
}
if sent != nil && sent.After(at) {
at = *sent
}
return at
}
if ok, on := applied(m, since, running, reports); !ok {
if ok, on := appliedEach(m, sinceFor, running, reports); !ok {
waiting = append(waiting, fmt.Sprintf("%s on %s", m, strings.Join(on, ", ")))
}
}
@@ -791,7 +821,7 @@ func planLine(p inventory.Plan, now time.Time) string {
case inventory.PlanSuperseded:
return fmt.Sprintf("%s %s %s", p.Repository, short(p.Commit), p.Note)
}
since := now.Sub(p.Updated).Round(time.Minute)
since := now.Sub(p.Updated).Round(time.Second)
late := ""
if since > planWaitBound {
late = " — LATE"
+31
View File
@@ -181,3 +181,34 @@ func TestAPlanSettlesAnAskedBuildFromTheRecords(t *testing.T) {
t.Errorf("a recorded failure did not fail the plan: %+v %+v", q, q.Modules["x"])
}
}
// The first machine's report, made between the send to it and the send to the rest, opens the gate for it:
// each machine is judged from its own send, not from the last one (novox/hq issue 256).
func TestTheGateJudgesEachMachineFromItsOwnSend(t *testing.T) {
built := time.Date(2026, 10, 5, 18, 22, 0, 0, time.UTC)
firstSent := built.Add(31 * time.Second)
firstReported := built.Add(43 * time.Second)
restSent := built.Add(58 * time.Second)
restReported := built.Add(74 * time.Second)
reports := []inventory.Reported{
{Node: "ace", At: &firstReported},
{Node: "g14", At: &restReported},
}
since := func(node string) time.Time {
if node == "ace" {
return firstSent
}
return restSent
}
if ok, waiting := appliedEach("build-agent", since, []string{"ace", "g14"}, reports); !ok {
t.Fatalf("the gate still waits on %v, though each reported after its own send", waiting)
}
// The old reading, every machine from the last send, is what held the plan.
if ok, _ := applied("build-agent", restSent, []string{"ace", "g14"}, reports); ok {
t.Fatal("the single-moment reading should hold the first machine back")
}
early := built.Add(10 * time.Second)
if ok, waiting := appliedEach("build-agent", since, []string{"ace"}, []inventory.Reported{{Node: "ace", At: &early}}); ok || waiting[0] != "ace" {
t.Fatal("a report from before the machine was sent opened the gate")
}
}