feat(jobs): expose job heartbeat and scan backlog

新增 JobActivity 持有作业心跳与扫描积压，JobRunner 每轮 tick 上报（失败只计数、不刷新心跳），PipelineMetrics 注册 msgx.pipeline.job.heartbeat_age_seconds / ticks.total / failures.total / last_sweep_selected，新增 JobRunnerHealthIndicator（未启动 UP 并标 started=false；已启动但心跳缺失或超过 3×扫描周期 DOWN）。对应 reference 作业健康段与 [G-JOB-HEARTBEAT]。

验证：./gradlew test 145 tests / 0 fail / 0 skipped（新增 JobRunnerTest 2、JobRunnerHealthIndicatorTest 4、PipelineMetrics 指标注册 1）。

src/test/kotlin/com/gzzn/omms/msgexchange/infra/health/JobRunnerHealthIndicatorTest.kt

@@ -0,0 +1,70 @@
+package com.gzzn.omms.msgexchange.infra.health
+
+import com.gzzn.omms.msgexchange.MutableClock
+import com.gzzn.omms.msgexchange.infra.metrics.JobActivity
+import io.micronaut.health.HealthStatus
+import org.junit.jupiter.api.Assertions.assertEquals
+import org.junit.jupiter.api.Test
+
+/**
+ * 作业存活判定：没启动不误报；启动了但心跳缺失或超过 3 个扫描周期必须 DOWN；
+ * 正常心跳保持 UP 并把扫描积压等事实带进详情。
+ */
+class JobRunnerHealthIndicatorTest {
+
+    private val now = MutableClock.BASE
+
+    private fun details(result: io.micronaut.management.health.indicator.HealthResult): Map<String, Any> {
+        @Suppress("UNCHECKED_CAST")
+        return result.details as Map<String, Any>
+    }
+
+    @Test
+    fun `a job runner that never started is up but flagged`() {
+        val result = jobRunnerHealth(JobActivity().snapshot(), now)
+
+        assertEquals(HealthStatus.UP, result.status)
+        assertEquals(false, details(result)["started"])
+        assertEquals(-1L, details(result)["lastTickAgeSeconds"])
+    }
+
+    @Test
+    fun `a started job with a fresh heartbeat is up`() {
+        val activity = JobActivity().apply {
+            started()
+            tickFinished(now.minusSeconds(30), durationMillis = 12, sweepSelected = 3)
+        }
+
+        val result = jobRunnerHealth(activity.snapshot(), now)
+
+        assertEquals(HealthStatus.UP, result.status)
+        assertEquals(1L, details(result)["ticks"])
+        assertEquals(3, details(result)["lastSweepSelected"])
+        assertEquals(12L, details(result)["lastTickDurationMs"])
+    }
+
+    @Test
+    fun `a started job that never ticked or went stale is down`() {
+        val neverTicked = JobActivity().apply { started() }
+        assertEquals(HealthStatus.DOWN, jobRunnerHealth(neverTicked.snapshot(), now).status)
+
+        val stale = JobActivity().apply {
+            started()
+            tickFinished(now.minusSeconds(STALE_AFTER_SECONDS + 1), durationMillis = 5, sweepSelected = 0)
+        }
+        assertEquals(HealthStatus.DOWN, jobRunnerHealth(stale.snapshot(), now).status)
+    }
+
+    @Test
+    fun `a failed tick counts up without refreshing the heartbeat`() {
+        val activity = JobActivity().apply { started() }
+
+        activity.tickFailed()
+        activity.tickFailed()
+
+        val snapshot = activity.snapshot()
+        assertEquals(2L, snapshot.failures)
+        assertEquals(0L, snapshot.ticks)
+        assertEquals(null, snapshot.lastTickAt)
+    }
+}

src/test/kotlin/com/gzzn/omms/msgexchange/infra/metrics/PipelineMetricsTest.kt

@@ -58,6 +58,20 @@ class PipelineMetricsTest {
         assertEquals(before + 2.0, gauge("msgx.pipeline.hole.aged_out.total"), 0.001)
     }
 
+    @Test
+    fun `job heartbeat gauges are registered and reflect job activity`() {
+        val activity = ctx.getBean(JobActivity::class.java)
+        val ticksBefore = gauge("msgx.pipeline.job.ticks.total")
+        val failuresBefore = gauge("msgx.pipeline.job.failures.total")
+
+        activity.tickFinished(Instant.now(), durationMillis = 12, sweepSelected = 7)
+
+        assertEquals(ticksBefore + 1.0, gauge("msgx.pipeline.job.ticks.total"), 0.001)
+        assertEquals(failuresBefore, gauge("msgx.pipeline.job.failures.total"), 0.001)
+        assertEquals(7.0, gauge("msgx.pipeline.job.last_sweep_selected"), 0.001)
+        org.junit.jupiter.api.Assertions.assertTrue(gauge("msgx.pipeline.job.heartbeat_age_seconds") >= 0.0)
+    }
+
     private fun gauge(name: String): Double =
         requireNotNull(registry.find(name).gauge()) { "gauge not registered: $name" }.value()
 }

src/test/kotlin/com/gzzn/omms/msgexchange/jobs/JobRunnerTest.kt

@@ -0,0 +1,75 @@
+package com.gzzn.omms.msgexchange.jobs
+
+import com.gzzn.omms.msgexchange.MutableClock
+import com.gzzn.omms.msgexchange.config.HistoryProps
+import com.gzzn.omms.msgexchange.config.MailboxProps
+import com.gzzn.omms.msgexchange.config.OperationDayProps
+import com.gzzn.omms.msgexchange.config.PipelineProps
+import com.gzzn.omms.msgexchange.domain.ProcStatus
+import com.gzzn.omms.msgexchange.infra.metrics.JobActivity
+import com.gzzn.omms.msgexchange.infra.persistence.BackfillDue
+import com.gzzn.omms.msgexchange.infra.persistence.ProcStateRepository
+import com.gzzn.omms.msgexchange.infra.stub.StubFlightState
+import com.gzzn.omms.msgexchange.infra.stub.StubInbox
+import com.gzzn.omms.msgexchange.infra.stub.StubMsgEvents
+import com.gzzn.omms.msgexchange.infra.stub.StubPipelineLock
+import com.gzzn.omms.msgexchange.infra.stub.StubPipelineTx
+import com.gzzn.omms.msgexchange.infra.stub.StubProcState
+import com.gzzn.omms.msgexchange.processing.BackfillService
+import com.gzzn.omms.msgexchange.processing.MessageLifecycleGate
+import org.junit.jupiter.api.Assertions.assertEquals
+import org.junit.jupiter.api.Assertions.assertNull
+import org.junit.jupiter.api.Test
+import java.time.Instant
+
+/**
+ * 作业心跳的接线：一轮 tick 跑完必须上报心跳与扫描积压；tick 抛错只累计失败、
+ * **不刷新心跳时刻**（停摆由心跳年龄暴露）。用 `TestClocks` 与内存适配器，不 sleep。
+ */
+class JobRunnerTest {
+
+    private val clock = MutableClock(MutableClock.BASE)
+    private val props = PipelineProps()
+
+    private fun historySweep() = HistorySweepJob(
+        StubFlightState(), StubMsgEvents(), HistoryProps().apply { historyStoreEnabled = false },
+        OperationDayProps(), StubPipelineTx(), StubPipelineLock(),
+    )
+
+    private fun runner(proc: ProcStateRepository, activity: JobActivity) = JobRunner(
+        BackfillService(proc, StubInbox(), MailboxProps(), props, clock, MessageLifecycleGate()),
+        historySweep(), clock, activity, OperationDayProps(),
+    )
+
+    @Test
+    fun `a completed tick reports the heartbeat and the backfill scan backlog`() {
+        val proc = StubProcState()
+        proc.insertIfAbsent(1L, MutableClock.BASE.minusSeconds(30L * 86400))
+        proc.markTerminal(1L, ProcStatus.SUCCEEDED, now = MutableClock.BASE)   // 终态未打标 → 本轮到期待办
+        val activity = JobActivity()
+
+        runner(proc, activity).tickOnce()
+
+        val snapshot = activity.snapshot()
+        assertEquals(1, snapshot.lastSweepSelected)
+        assertEquals(1L, snapshot.ticks)
+        assertEquals(0L, snapshot.failures)
+        assertEquals(MutableClock.BASE, snapshot.lastTickAt)
+    }
+
+    @Test
+    fun `a failing tick counts the failure and leaves the heartbeat untouched`() {
+        val broken = object : ProcStateRepository by StubProcState() {
+            override fun findBackfillDue(now: Instant, overdueBefore: Instant, limit: Int): List<BackfillDue> =
+                throw IllegalStateException("db-down")
+        }
+        val activity = JobActivity()
+
+        runner(broken, activity).tickOnce()
+
+        val snapshot = activity.snapshot()
+        assertNull(snapshot.lastTickAt)
+        assertEquals(0L, snapshot.ticks)
+        assertEquals(1L, snapshot.failures)
+    }
+}