Index: rmt.scm
==================================================================
--- rmt.scm
+++ rmt.scm
@@ -62,11 +62,11 @@
       (debug:print 0 *default-log-port* "ERROR: 15 tries to start/connect to server. Giving up.")
       (exit 1))
      ;; reset the connection if it has been unused too long
      ((and *runremote*
            (remote-conndat *runremote*)
-	   (let ((expire-time (- start-time (remote-server-timeout *runremote*))))
+	   (let ((expire-time (+ (- start-time (remote-server-timeout *runremote*))(random 30)))) ;; add 30 seconds of noise so that not all running tests expire at the same time causing a storm of server starts
 	     (< (http-transport:server-dat-get-last-access (remote-conndat *runremote*)) expire-time)))
       (debug:print-info 12 *default-log-port* "rmt:send-receive, case  8")
       (remote-conndat-set! *runremote* #f)
       (mutex-unlock! *rmt-mutex*)
       (rmt:send-receive cmd rid params attemptnum: attemptnum))

Index: server.scm
==================================================================
--- server.scm
+++ server.scm
@@ -221,11 +221,14 @@
 		     (mod-time   (list-ref rec 0)))
 		 ;; (print "start-time: " start-time " mod-time: " mod-time)
 		 (and start-time mod-time
 		      (> (- now start-time) 0)    ;; been running at least 0 seconds
 		      (< (- now mod-time)   16)   ;; still alive - file touched in last 16 seconds
-		      (< (- now start-time) (string->number (or (configf:lookup *configdat* "server" "runtime") "3600"))) ;; under one hour running time
+		      (< (- now start-time) 
+                         (+ (- (string->number (or (configf:lookup *configdat* "server" "runtime") "3600"))
+                               180)
+                            (random 360))) ;; under one hour running time +/- 180
 		      )))
 	     srvlst)
      (lambda (a b)
        (< (list-ref a 3)
 	  (list-ref b 3))))))