feat(watchdog): hardware error monitor + unified component status store
- Add platform/error_patterns.go: pluggable table of kernel log patterns (NVIDIA/GPU, PCIe AER, storage I/O, MCE, EDAC) — extend by adding one struct - Add app/component_status_db.go: persistent JSON store (component-status.json) keyed by "pcie:BDF", "storage:dev", "cpu:all", "memory:all"; OK never downgrades Warning or Critical - Add webui/kmsg_watcher.go: goroutine reads /dev/kmsg during SAT tasks, writes Warning to DB for matched hardware errors - Fix task status: overall_status=FAILED in summary.txt now marks task failed - Audit routine overlays component DB statuses into bee-audit.json on every read Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -173,13 +173,14 @@ func resolvePlatformStressPreset(profile string) platform.PlatformStressOptions
|
||||
|
||||
// taskQueue manages a priority-ordered list of tasks and runs them one at a time.
|
||||
type taskQueue struct {
|
||||
mu sync.Mutex
|
||||
tasks []*Task
|
||||
trigger chan struct{}
|
||||
opts *HandlerOptions // set by startWorker
|
||||
statePath string
|
||||
logsDir string
|
||||
started bool
|
||||
mu sync.Mutex
|
||||
tasks []*Task
|
||||
trigger chan struct{}
|
||||
opts *HandlerOptions // set by startWorker
|
||||
statePath string
|
||||
logsDir string
|
||||
started bool
|
||||
kmsgWatcher *kmsgWatcher
|
||||
}
|
||||
|
||||
var globalQueue = &taskQueue{trigger: make(chan struct{}, 1)}
|
||||
@@ -411,8 +412,16 @@ func (q *taskQueue) worker() {
|
||||
q.persistLocked()
|
||||
q.mu.Unlock()
|
||||
|
||||
if q.kmsgWatcher != nil && isSATTarget(t.Target) {
|
||||
q.kmsgWatcher.NotifyTaskStarted(t.ID, t.Target)
|
||||
}
|
||||
|
||||
q.runTask(t, j, ctx)
|
||||
|
||||
if q.kmsgWatcher != nil {
|
||||
q.kmsgWatcher.NotifyTaskFinished(t.ID)
|
||||
}
|
||||
|
||||
q.mu.Lock()
|
||||
now2 := time.Now()
|
||||
t.DoneAt = &now2
|
||||
@@ -618,6 +627,19 @@ func (q *taskQueue) runTask(t *Task, j *jobState, ctx context.Context) {
|
||||
return
|
||||
}
|
||||
|
||||
// If the SAT archive was produced, check overall_status and write to component DB.
|
||||
if archive != "" {
|
||||
archivePath := app.ExtractArchivePath(archive)
|
||||
if err == nil {
|
||||
if app.ReadSATOverallStatus(archivePath) == "FAILED" {
|
||||
err = fmt.Errorf("SAT overall_status=FAILED (see summary.txt)")
|
||||
}
|
||||
}
|
||||
if db := q.statusDB(); db != nil {
|
||||
app.ApplySATResultToDB(db, t.Target, archivePath)
|
||||
}
|
||||
}
|
||||
|
||||
if err != nil {
|
||||
if ctx.Err() != nil {
|
||||
j.append("Aborted.")
|
||||
@@ -634,6 +656,13 @@ func (q *taskQueue) runTask(t *Task, j *jobState, ctx context.Context) {
|
||||
j.finish("")
|
||||
}
|
||||
|
||||
func (q *taskQueue) statusDB() *app.ComponentStatusDB {
|
||||
if q.opts == nil || q.opts.App == nil {
|
||||
return nil
|
||||
}
|
||||
return q.opts.App.StatusDB
|
||||
}
|
||||
|
||||
func splitLines(s string) []string {
|
||||
var out []string
|
||||
for _, l := range splitNL(s) {
|
||||
|
||||
Reference in New Issue
Block a user