fix: add refreshing errors to the retrieval state:

- add alerts about refreshing errors to the state of the instance (the `retrieving` section) 
- change the output level of data refreshing errors from `debug` to `error`

Author: agneum

engine/internal/retrieval/engine/postgres/logical/dump.go

@@ -481,7 +481,7 @@ func (d *DumpJob) cleanupDumpLocation(ctx context.Context, dumpContID string, db
 		Tty: true,
 		Cmd: cleanupCmd,
 	}); err != nil {
-		log.Dbg(out)
+		log.Err(out)
 		return errors.Wrap(err, "failed to clean up dump location")
 	}
 
@@ -511,8 +511,9 @@ func (d *DumpJob) dumpDatabase(ctx context.Context, dumpContID, dbName string, d
 		Cmd: dumpCommand,
 		Env: d.getExecEnvironmentVariables(),
 	}); err != nil {
-		log.Dbg(output)
-		return errors.Wrap(err, "failed to dump a database")
+		log.Err("Dump command failed: ", output)
+
+		return fmt.Errorf("failed to dump a database: %w. Output: %s", err, output)
 	}
 
 	log.Msg(fmt.Sprintf("Dumping job for the database %q has been finished", dbName))

engine/internal/retrieval/engine/postgres/logical/restore.go

@@ -519,12 +519,14 @@ func (r *RestoreJob) restoreDB(ctx context.Context, contID, dbName string, dbDef
 		Env: []string{"PGAPPNAME=" + dleRetrieval},
 	})
 
-	if output != "" {
-		log.Dbg("Output of the restore command: ", output)
+	if err != nil {
+		log.Err("Restore command failed: ", output)
+
+		return fmt.Errorf("failed to exec restore command: %w. Output: %s", err, output)
 	}
 
-	if err != nil {
-		return errors.Wrap(err, "failed to exec restore command")
+	if output != "" {
+		log.Dbg("Output of the restore command: ", output)
 	}
 
 	if err := r.defineDSA(ctx, dbDefinition, contID, dbName); err != nil {

engine/internal/retrieval/retrieval.go

@@ -222,10 +222,10 @@ func (r *Retrieval) Run(ctx context.Context) error {
 	}
 
 	if err := r.run(runCtx, fsManager); err != nil {
-		alert := telemetry.Alert{Level: models.RefreshFailed,
-			Message: fmt.Sprintf("Failed to perform initial data retrieving: %s", r.State.Mode)}
-		r.State.addAlert(alert)
-		r.tm.SendEvent(ctx, telemetry.AlertEvent, alert)
+		r.State.addAlert(telemetry.Alert{Level: models.RefreshFailed, Message: err.Error()})
+		// Build a generic message to avoid sending sensitive data.
+		r.tm.SendEvent(ctx, telemetry.AlertEvent, telemetry.Alert{Level: models.RefreshFailed,
+			Message: fmt.Sprintf("Failed to perform initial data retrieving: %s", r.State.Mode)})
 
 		return err
 	}
@@ -382,6 +382,10 @@ func (r *Retrieval) RefreshData(ctx context.Context, poolName string) error {
 
 		if err != nil {
 			r.State.Status = models.Failed
+			r.State.addAlert(telemetry.Alert{
+				Level:   models.RefreshFailed,
+				Message: err.Error(),
+			})
 
 			fsm.Pool().SetStatus(resources.EmptyPool)
 		}
@@ -436,6 +440,10 @@ func (r *Retrieval) SnapshotData(ctx context.Context, poolName string) error {
 
 		if err != nil {
 			r.State.Status = models.Failed
+			r.State.addAlert(telemetry.Alert{
+				Level:   models.RefreshFailed,
+				Message: err.Error(),
+			})
 
 			fsm.Pool().SetStatus(resources.EmptyPool)
 		}
@@ -543,10 +551,10 @@ func (r *Retrieval) setupScheduler(ctx context.Context) {
 func (r *Retrieval) refreshFunc(ctx context.Context) func() {
 	return func() {
 		if err := r.FullRefresh(ctx); err != nil {
-			alert := telemetry.Alert{Level: models.RefreshFailed, Message: "Failed to run full-refresh"}
+			alert := telemetry.Alert{Level: models.RefreshFailed, Message: err.Error()}
 			r.State.addAlert(alert)
-			r.tm.SendEvent(ctx, telemetry.AlertEvent, alert)
-			log.Err(alert.Message, err)
+			r.tm.SendEvent(ctx, telemetry.AlertEvent, telemetry.Alert{Level: models.RefreshFailed, Message: "Failed to run full-refresh"})
+			log.Err(alert.Message)
 		}
 	}
 }