Megatest

				   cmd-in
				   (string->symbol cmd-in)))
            (params            (vector-ref dat 1))
            (start-t           (current-milliseconds))
            (readonly-mode     (dbr:dbstruct-read-only dbstruct))
            (readonly-command  (member cmd api:read-only-queries))
            (writecmd-in-readonly-mode (and readonly-mode (not readonly-command)))
            #;(foo               (begin
                                 (common:telemetry-log (conc "api-in:"(->string cmd))
                                                       payload: `((params . ,params)))
                                 
                                 #t))
            (res    
             (if writecmd-in-readonly-mode
                 (conc "attempt to run write command "cmd" on a read-only database")

       ;; save all stats
       (let ((delta-t (- (current-milliseconds)
			 start-t)))
	 (hash-table-set! *db-api-call-time* cmd
			  (cons delta-t (hash-table-ref/default *db-api-call-time* cmd '()))))
       (if writecmd-in-readonly-mode
           (begin
             #;(common:telemetry-log (conc "api-out:"(->string cmd))
                                   payload: `((params . ,params)
                                              (ok-res . #t)))
	     (vector #f res))
           (begin
             #;(common:telemetry-log (conc "api-out:"(->string cmd))
                                   payload: `((params . ,params)
                                              (ok-res . #f)))
             (vector #t res))))))))

;; http-server  send-response
;;                 api:process-request
;;                    db:*

	     (begin
	       (debug:print 0 *default-log-port* "joining threads failed. exn=" exn)
               #t) ;; just ignore it, it might have died in the meantime so joining it will throw an exception
             (thread-join! thread))
           )))
   (hash-table-keys *common:thread-punchlist*)))

#;(define *common:telemetry-log-state* 'startup)
#;(define *common:telemetry-log-socket* #f)

#;(define (common:telemetry-log-open)
  (if (eq? *common:telemetry-log-state* 'startup)
      (let* ((serverhost (configf:lookup *configdat* "telemetry" "host"))
             (serverport (configf:lookup-number *configdat* "telemetry" "port"))
             (user (or (get-environment-variable "USER") "unknown"))
             (host (or (get-environment-variable "HOST") "unknown")))
        (set! *common:telemetry-log-state*
              (handle-exceptions
               exn
               (begin
                 (debug:print-info 0 *default-log-port* "common-telemetry-log open udp port failure")
                 'broken)
               (if (and serverhost serverport user host)
                   (let* ((s (udp-open-socket)))
                     ;;(udp-bind! s #f 0)
                     (udp-connect! s serverhost serverport)
                     (set! *common:telemetry-log-socket* s)
                     'open)
                   'not-needed))))))
  
#;(define (common:telemetry-log event #!key (payload '()))
  (if (eq? *common:telemetry-log-state* 'startup)
      (common:telemetry-log-open))

  (if (eq? 'open *common:telemetry-log-state*)
      (handle-exceptions
       exn
       (begin

                   (base64:base64-encode
                    (z3:encode-buffer
                     (with-output-to-string (lambda () (pp payload))))))
                  (msg     (conc user":"host":"start":"(current-process-id)":"(car (argv))":"
                                 toppath":"payload-serialized)))
             (udp-send *common:telemetry-log-socket* msg))))))
  
#;(define (common:telemetry-log-close)
  (when (or (member *common:telemetry-log-state* '(broken-or-no-server-preclose open)) *common:telemetry-log-socket*)
    (handle-exceptions
     exn
     (begin
       (define *common:telemetry-log-state* 'closed-fail)
       (debug:print-info 0 *default-log-port* "common-telemetry-log closure failure")
       )
     (begin
       (define *common:telemetry-log-state* 'closed)
       (udp-close-socket *common:telemetry-log-socket*)
       (set! *common:telemetry-log-socket* #f)))))

;;    ;; 	(db:general-call dbdat 'state-status-msg (list state status msg test-id))
;;    ;; 	(db:general-call dbdat 'state-status     (list state status test-id)))
;;    (db:set-state-status-and-roll-up-items dbstruct run-id test-id #f state status msg)
;;    ;; process the test_data table
;;    (if (and test-id state status (equal? status "AUTO")) 
;; 	(db:test-data-rollup dbstruct run-id test-id status))
;;    (mt:process-triggers dbstruct run-id test-id state status)))

;; NOT FINISHED 
(define (db:calc-state-status-toplevel state status tl-state tl-status)
  `(,state ,status))

;;   (match state
;;     (("COMPLETED")
;;      (match `(,tl-state ,tl-status)
;;        (("COMPLETED" "PASS") `(,state ,status))
;;        (("COMPLETED" thestatus)
;; 	(case (string->symbol thestatus)
;; 	  ((ABORT CHECK DEAD)
;; 	   (if `("COMPLETED" ,thestatus))
;; 	(match `(,thestatus ,status)
;; 	  (("FAIL" "ABORT") '("COMPLETED" "ABORT"))
;; 	  (("FAIL" "CHECK") '("COMPLETED" "CHECK"))
;; 	  (("FAIL" "DEAD")  '("COMPLETED" "DEAD"))
;; 	  (("WARN" "FAIL")  '("COMPLETED" "FAIL"))
;; 	  (("WARN" "CHECK") '("COMPLETED" "CHECK"))
;; 	  (("WARN" "DEAD")
       

(define (db:set-state-status-and-roll-up-items-fast dbstruct run-id test-name item-path state status comment)
  ;; establish info on incoming test followed by info on top level test
  ;; BBnote - for mode itemwait, linkage between upstream test & matching item status is propagated to run queue in db:prereqs-not-met
  (mutex-lock! *db-transaction-mutex*)
  (db:with-db
   dbstruct #f #f
   (lambda (db)
     (let ((tr-res
	    (sqlite3:with-transaction
	     db
	     (lambda ()
	       (let* ((testdat      (if (number? test-name)
					(db:get-test-info-by-id dbstruct run-id test-name) ;; test-name is actually a test-id
					(db:get-test-info       dbstruct run-id test-name item-path)))
		      (test-id      (db:test-get-id testdat))
		      (test-name    (if (number? test-name)
					(db:test-get-testname testdat)
					test-name))
		      (item-path    (db:test-get-item-path testdat))
		      (tl-testdat   (db:get-test-info dbstruct run-id test-name ""))
		      (tl-test-id   (if tl-testdat
					(db:test-get-id tl-testdat)
					#f)))
		 (if (member state '("LAUNCHED" "REMOTEHOSTSTART"))
		     (db:general-call dbstruct 'set-test-start-time (list test-id)))
		 (if (member state '("LAUNCHED" "REMOTEHOSTSTART" "RUNNING"))
		     (begin ;; set test to launched and top level to RUNNING and we are done
		       (if tl-test-id
			   (db:test-set-state-status db run-id tl-test-id "RUNNING" #f #f)) ;; we are still in the transaction - must access the db and not the dbstruct
		       (db:test-set-state-status db run-id test-id state status #f))
		     (begin ;; 
		       (db:test-set-state-status db run-id test-id state status comment) ;; this call sets the item state/status
		       (if (not (equal? item-path "")) ;; only roll up IF incoming test is an item
			   (let* ((tl-state  (db:test-get-state tl-testdat))
				  (tl-status (db:test-get-status tl-testdat))
				  (newss     (db:calc-state-status-toplevel state status tl-state tl-status)))
			     (match newss
			       ((new-tl-state new-tl-status)
				(db:test-set-state-status db run-id tl-test-id new-tl-state new-tl-status))
			       (else
				#f)))))))))))
       (mutex-unlock! *db-transaction-mutex*)
       (if (and test-id state status (equal? status "AUTO")) 
	   (db:test-data-rollup dbstruct run-id test-id status))
       tr-res))))


;; state is the priority rollup of all states
;; status is the priority rollup of all completed statesfu
;;
;; if test-name is an integer work off that instead of test-name test-path
;;
(define (db:set-state-status-and-roll-up-items dbstruct run-id test-name item-path state status comment)

              (sqlite3:with-transaction
               db
               (lambda ()
                 ;; NB// Pass the db so it is part fo the transaction
                 (db:test-set-state-status db run-id test-id state status comment) ;; this call sets the item state/status
                 (if (not (equal? item-path "")) ;; only roll up IF incoming test is an item
                     (let* ((state-status-counts  (db:get-all-state-status-counts-for-test dbstruct run-id test-name item-path state status)) ;; item-path is used to exclude current state/status of THIS test
			    (state-stauses        (db:roll-up-rules state-status-counts state status))
                          (newstate (car state-stauses))
                          (newstatus (cadr state-stauses)))
                       (debug:print 4 *default-log-port* "BB> tl-test-id="tl-test-id" ; "test-name":"item-path" newstate="newstate" newstatus="newstatus" len(sscs)="(length state-status-counts)  " state-status-counts: "
							(apply conc
                  (map (lambda (x)
                     (conc
                     		(with-output-to-string (lambda () (pp (dbr:counts->alist x)))) " | "))

    ;; (tests:set-full-meta-info test-id run-id (calc-minutes) work-area)
    (tests:set-full-meta-info #f test-id run-id (calc-minutes) work-area 10)

    (let loop ((minutes   (calc-minutes))
	       (cpu-load  (alist-ref 'adj-core-load (common:get-normalized-cpu-load #f)))
	       (disk-free (get-df (current-directory)))
               (last-sync (current-seconds)))
      ;; (common:telemetry-log "zombie" (conc "launch:monitor-job - top of loop encountered at "(current-seconds)" with last-sync="last-sync))
      (let* ((over-time     (> (current-seconds) (+ last-sync update-period)))
             (new-cpu-load  (let* ((load  (alist-ref 'adj-core-load (common:get-normalized-cpu-load #f)))
                                   (delta (abs (- load cpu-load))))
                              (if (> delta 0.1) ;; don't bother updating with small changes
                                  load
                                  #f)))
             (new-disk-free (let* ((df    (if over-time ;; only get df every 30 seconds

             (do-sync       (or new-cpu-load new-disk-free over-time))

             (test-info   (rmt:get-test-info-by-id run-id test-id))
             (state       (db:test-get-state test-info))
             (status      (db:test-get-status test-info))
             (kill-reason  "no kill reason specified")
             (kill-job?    #f))
        ;; (common:telemetry-log "zombie" (conc "launch:monitor-job - decision time encountered at "(current-seconds)" with last-sync="last-sync" do-sync="do-sync" over-time="over-time" update-period="update-period))
        (cond
         ((test-get-kill-request run-id test-id)
          (set! kill-reason "KILLING TEST since received kill request (KILLREQ)")
          (set! kill-job? #t))
         ((and runtlim (> (- (current-seconds) start-seconds) runtlim))
          (set! kill-reason (conc "KILLING TEST DUE TO TIME LIMIT EXCEEDED! Runtime=" (- (current-seconds) start-seconds) " seconds, limit=" runtlim))
          (set! kill-job? #t))
         ((equal? status "DEAD")
          (tests:update-central-meta-info run-id test-id new-cpu-load new-disk-free (calc-minutes) #f #f)
          (rmt:set-state-status-and-roll-up-items run-id test-id 'foo "RUNNING" "n/a" "was marked dead; really still running.")
          ;;(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 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
	      ;;       section and the runit section? Or add a loop that tries three times with a 1/4 second

		   (if last-time
		       (< (current-seconds)(+ last-time 5))
		       #f))))
    (if useres
	(let ((result (vector-ref res 1)))
	  (debug:print 4 *default-log-port* "Using lazy value res: " result)
	  result)
	(let ((newres (rmt:get-prereqs-not-met run-id waitons ref-item-path mode itemmaps)))
	  (hash-table-set! *pre-reqs-met-cache* key (vector (current-seconds) newres))
	  newres))))

(define (mt:get-run-stats dbstruct run-id)
;;  Get run stats from local access, move this ... but where?
  (db:get-run-stats dbstruct run-id))

(define (rmt:test-get-paths-matching-keynames-target-new keynames target res testpatt statepatt statuspatt runname)
  (let ((run-ids (rmt:get-run-ids-matching-target keynames target res runname testpatt statepatt statuspatt)))
    (apply append 
	   (map (lambda (run-id)
		  (rmt:send-receive 'test-get-paths-matching-keynames-target-new run-id (list run-id keynames target res testpatt statepatt statuspatt runname)))
	   run-ids))))

;; NOTE: rmt functions can NEVER have key params as they might be called as local
(define (rmt:get-prereqs-not-met run-id waitons ref-test-name ref-item-path mode itemmaps) ;; #!key (mode '(normal))(itemmaps #f))
  (rmt:send-receive 'get-prereqs-not-met run-id (list run-id waitons ref-test-name ref-item-path mode itemmaps)))

(define (rmt:get-count-tests-running-for-run-id run-id fastmode)
  (rmt:send-receive 'get-count-tests-running-for-run-id run-id (list run-id fastmode)))

(define (rmt:get-not-completed-cnt run-id)
  (rmt:send-receive 'get-not-completed-cnt run-id (list run-id)))

;;    => review of a previously seen test is higher priority of never visited test
;; reg - list of previously visited tests
;; tal - list of never visited tests
;;   prefer next hed to be from reg than tal.

(define runs:nothing-left-in-queue-count 0)

(define (runs:lazy-get-prereqs-not-met  testdat run-id waitons hed item-path #!key (mode '(normal))(itemmaps #f)) ;; mode: testmode itemmaps: itemmaps)
  (if (and (runs:testdat-prereqs-not-met testdat)
	   (< (- (current-seconds) (runs:testdat-last-update testdat)) 10)) ;; only refresh for this test if it has been at least 10 seconds
      (runs:testdat-prereqs-not-met testdat)
      (let* ((res (let ((res (rmt:get-prereqs-not-met run-id waitons hed item-path mode itemmaps)))
		    (if (list? res)
			res
			(begin
			  (debug:print 0 *default-log-port*
				       "ERROR: rmt:get-prereqs-not-met returned non-list!\n"
				       "  res=" res " run-id=" run-id " waitons=" waitons " hed=" hed " item-path=" item-path " testmode=" mode " itemmaps=" itemmaps)
			  '())))))

;;	         (tal         (cdr sorted-test-names))
;;	         (reg         '()) ;; registered, put these at the head of tal 
;;	         (reruns      '()))
(define (runs:expand-items hed tal reg reruns regfull newtal jobgroup max-concurrent-jobs run-id waitons item-path testmode test-record
			   can-run-more items runname tconfig reglen test-registry test-records itemmaps testdat)
  (let* ((loop-list       (list hed tal reg reruns))
	 (prereqs-not-met (runs:lazy-get-prereqs-not-met testdat run-id waitons hed item-path mode: testmode itemmaps: itemmaps))








	 (have-itemized   (not (null? (lset-intersection eq? testmode '(itemmatch itemwait)))))

	 (fails           (runs:calc-fails prereqs-not-met))
	 (prereq-fails    (runs:calc-prereq-fail prereqs-not-met))
	 (non-completed   (runs:calc-not-completed prereqs-not-met))
	 (runnables       (runs:calc-runnable prereqs-not-met))
         (unexpanded-prereqs
          (filter (lambda (testname)
                    (let* ((test-rec (hash-table-ref test-records testname))

	 (run-limits-info        (runs:dat-can-run-more-tests runsdat))
	 ;; (runs:can-run-more-tests run-id jobgroup max-concurrent-jobs)) ;; look at the test jobgroup and tot jobs running
	 (have-resources         (car run-limits-info))
	 (num-running            (list-ref run-limits-info 1))
	 (num-running-in-jobgroup(list-ref run-limits-info 2)) 
	 (max-concurrent-jobs    (list-ref run-limits-info 3))
	 (job-group-limit        (list-ref run-limits-info 4))


	 (fails                  (if (list? prereqs-not-met) ;; TODO: rename fails to failed-prereqs
				      (runs:calc-fails prereqs-not-met)
				      (begin
					(debug:print-error 0 *default-log-port* "prereqs-not-met is not a list! " prereqs-not-met)
					'())))
	 (non-completed           (filter (lambda (x)             ;; remove hed from not completed list, duh, of course it is not completed!
					    (not (equal? x hed)))

                   registry-mutex: registry-mutex
                   flags: flags
                   keyvals: keyvals
                   run-info: run-info
                   ;; newtal: newtal
                   all-tests-registry: all-tests-registry
                   ;; itemmaps: itemmaps

                   ;; can-run-more-tests: (runs:can-run-more-tests run-id jobgroup max-concurrent-jobs) ;; look at the test jobgroup and tot jobs running
                   )))

    ;; Initialize the test-registery hash with tests that already have a record
    ;; convert state to symbol and use that as the hash value
    (for-each (lambda (trec)
		(let ((id (db:test-get-id        trec))

;;======================================================================

;;======================================================================
;; Tests
;;======================================================================

(declare (unit tests))
;;(declare (uses lock-queue))
(declare (uses db))
(declare (uses tdb))
(declare (uses common))
;; (declare (uses dcommon)) ;; needed for the steps processing
(declare (uses items))
(declare (uses runconfig))
;; (declare (uses sdb))