@@ -251,17 +251,12 @@
           ;;(set! kill-reason "KILLING TEST because it was marked as DEAD by launch:handle-zombie-tests (might indicate really overloaded server or else overzealous setup.deadtime)") ;; MARK RUNNING
           (set! kill-job? #f)))
 
         (debug:print 4 *default-log-port* "cpu: " new-cpu-load " disk: " new-disk-free " last-sync: " last-sync " do-sync: " do-sync)
         (launch:handle-zombie-tests run-id)
-        (when do-sync
-          ;;(with-output-to-file (conc (getenv "MT_TEST_RUN_DIR") "/last-loadinfo.log" #:append)
-          ;;  (lambda () (pp (list (current-seconds) new-cpu-load new-disk-free (calc-minutes)))))
-          (common:telemetry-log "zombie" (conc  "launch:monitor-job - dosync started at "(current-seconds)))
-          (tests:update-central-meta-info run-id test-id new-cpu-load new-disk-free (calc-minutes) #f #f)
-          (common:telemetry-log "zombie" (conc "launch:monitor-job - dosync finished at "(current-seconds))))
-        
+        (if do-sync ;; save meta data about the running of this test
+	    (tests:update-central-meta-info run-id test-id new-cpu-load new-disk-free (calc-minutes) #f #f))
 	(if kill-job? 
 	    (begin
               (debug:print-info 0 *default-log-port* "proceeding to kill test: "kill-reason)
 	      (mutex-lock! m)
 	      ;; NOTE: The pid can change as different steps are run. Do we need handshaking between this