Megatest

Check-in [447b257e2f]
Login
Overview
Comment:Limit to 10 dbs. Fully turned off running find-and-mark in runs.scm
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | v1.71
Files: files | file ages | folders
SHA1: 447b257e2fd0fdb2ea265807bfffa39e96bf93a4
User & Date: matt on 2023-05-15 18:59:39
Other Links: branch diff | manifest | tags
Context
2023-05-16
18:07
Changed version to 1.7015 check-in: a01734df49 user: mmgraham tags: v1.71, v1.7015
2023-05-15
18:59
Limit to 10 dbs. Fully turned off running find-and-mark in runs.scm check-in: 447b257e2f user: matt tags: v1.71
2023-05-08
16:46
Added with-transaction around a for-each-row check-in: eb2060809c user: mmgraham tags: v1.70
Changes

Modified db.scm from [d4dcfd72a3] to [ea93a4142c].

4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
         (changed_run_dbs (db:get-changed-run-ids since-time)) ;; gets the rundb numbers
         (all_run_ids 
          (db:with-db dbstruct #f #f 
            (lambda (dbdat db)
              (sqlite3:fold-row backcons '() db "SELECT id FROM runs"))
          )
         )
         (changed_run_ids (filter (lambda (run) (member (modulo run 100) changed_run_dbs)) all_run_ids))
         ;; TODO: couldn't we just use changed_run_ids for run_ids?
         (run_ids 
          (db:with-db dbstruct #f #f 
            (lambda (dbdat db)
              (sqlite3:fold-row backcons '() db "SELECT id FROM runs  WHERE last_update>=?" since-time))
          )
         )







|







4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
         (changed_run_dbs (db:get-changed-run-ids since-time)) ;; gets the rundb numbers
         (all_run_ids 
          (db:with-db dbstruct #f #f 
            (lambda (dbdat db)
              (sqlite3:fold-row backcons '() db "SELECT id FROM runs"))
          )
         )
         (changed_run_ids (filter (lambda (run) (member (modulo run (num-run-dbs)) changed_run_dbs)) all_run_ids))
         ;; TODO: couldn't we just use changed_run_ids for run_ids?
         (run_ids 
          (db:with-db dbstruct #f #f 
            (lambda (dbdat db)
              (sqlite3:fold-row backcons '() db "SELECT id FROM runs  WHERE last_update>=?" since-time))
          )
         )

Modified dbfile.scm from [a0594e55e5] to [8e0e70bc74].

40
41
42
43
44
45
46




47
48
49
50
51
52
53
	ports

	commonmod
	)

;; (import debugprint)





;;======================================================================
;;  R E C O R D S
;;======================================================================

;; a single Megatest area with it's multiple dbs is
;; managed in a dbstruct
;;







>
>
>
>







40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
	ports

	commonmod
	)

;; (import debugprint)

;; Parameters

(define num-run-dbs           (make-parameter 10))

;;======================================================================
;;  R E C O R D S
;;======================================================================

;; a single Megatest area with it's multiple dbs is
;; managed in a dbstruct
;;
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208

(define (db:dbname->path apath dbname)
  (conc apath"/"dbname))

;; POTENTIAL BUG: this implementation could produce a db file if run-id is neither #f or a number
(define (dbfile:run-id->dbname run-id)
  (cond
   ((number? run-id) (conc ".megatest/" (modulo run-id 100) ".db"))
   ((not run-id)     (conc ".megatest/main.db"))
   (else             run-id)))

;; Make the dbstruct, setup up auxillary db's and call for main db at least once
;;
;; called in http-transport and replicated in rmt.scm for *local* access. 
;;







|







198
199
200
201
202
203
204
205
206
207
208
209
210
211
212

(define (db:dbname->path apath dbname)
  (conc apath"/"dbname))

;; POTENTIAL BUG: this implementation could produce a db file if run-id is neither #f or a number
(define (dbfile:run-id->dbname run-id)
  (cond
   ((number? run-id) (conc ".megatest/" (modulo run-id (num-run-dbs)) ".db"))
   ((not run-id)     (conc ".megatest/main.db"))
   (else             run-id)))

;; Make the dbstruct, setup up auxillary db's and call for main db at least once
;;
;; called in http-transport and replicated in rmt.scm for *local* access. 
;;

Modified runs.scm from [6164fa6d2a] to [0a3622cb0c].

808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
		;; 				exn
		;; 				(begin
		;; 				  (print-call-chain)
		;; 				  (print " message: " ((condition-property-accessor 'exn 'message) exn)))
		;; 			      (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests
		;; 						    (any->number reglen) all-tests-registry)))
		;; 			  "runs:run-tests-queue"))
		 (th2        (make-thread (lambda ()			 ;; BBQ: why are we visiting ALL runs here?	    
					    ;; (rmt:find-and-mark-incomplete-all-runs))))) CAN'T INTERRUPT IT ...
					    (let ((run-ids (rmt:get-all-run-ids)))
					      (for-each (lambda (run-id)
							  (if keep-going
							      (handle-exceptions
							       exn
							       (debug:print 0 *default-log-port* "error in calling find-and-mark-incomplete for run-id " run-id ", exn=" exn)
							       (rmt:find-and-mark-incomplete run-id #f)))) ;; ovr-deadtime))) ;; could be root of https://hsdes.intel.com/appstore/article/#/220546828/main -- Title: Megatest jobs show DEAD even though they are still running (1.64/27)
							run-ids)))
					  "runs: mark-incompletes")))
	    ;; (thread-start! th1)
	    (thread-start! th2)
	    ;; (thread-join! th1)
	    ;; just do the main stuff in the main thread
	    (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests
								    (any->number reglen) all-tests-registry)
	    (set! keep-going #f)
	    (thread-join! th2)
	    ;; if run-count > 0 call, set -preclean and -rerun STUCK/DEAD
	    (if (> run-count 0) ;; handle reruns
		(begin
		  (if (not (hash-table-ref/default flags "-preclean" #f))
		      (hash-table-set! flags "-preclean" #t))
		  (if (not (hash-table-ref/default flags "-rerun" #f))
		      (hash-table-set! flags "-rerun" "ABORT,STUCK/DEAD,n/a,ZERO_ITEMS"))







|











|





|







808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
		;; 				exn
		;; 				(begin
		;; 				  (print-call-chain)
		;; 				  (print " message: " ((condition-property-accessor 'exn 'message) exn)))
		;; 			      (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests
		;; 						    (any->number reglen) all-tests-registry)))
		;; 			  "runs:run-tests-queue"))
		 #;(th2        (make-thread (lambda ()			 ;; BBQ: why are we visiting ALL runs here?	    
					    ;; (rmt:find-and-mark-incomplete-all-runs))))) CAN'T INTERRUPT IT ...
					    (let ((run-ids (rmt:get-all-run-ids)))
					      (for-each (lambda (run-id)
							  (if keep-going
							      (handle-exceptions
							       exn
							       (debug:print 0 *default-log-port* "error in calling find-and-mark-incomplete for run-id " run-id ", exn=" exn)
							       (rmt:find-and-mark-incomplete run-id #f)))) ;; ovr-deadtime))) ;; could be root of https://hsdes.intel.com/appstore/article/#/220546828/main -- Title: Megatest jobs show DEAD even though they are still running (1.64/27)
							run-ids)))
					  "runs: mark-incompletes")))
	    ;; (thread-start! th1)
	    ;; (thread-start! th2)
	    ;; (thread-join! th1)
	    ;; just do the main stuff in the main thread
	    (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests
								    (any->number reglen) all-tests-registry)
	    (set! keep-going #f)
	    #;(thread-join! th2)
	    ;; if run-count > 0 call, set -preclean and -rerun STUCK/DEAD
	    (if (> run-count 0) ;; handle reruns
		(begin
		  (if (not (hash-table-ref/default flags "-preclean" #f))
		      (hash-table-set! flags "-preclean" #t))
		  (if (not (hash-table-ref/default flags "-rerun" #f))
		      (hash-table-set! flags "-rerun" "ABORT,STUCK/DEAD,n/a,ZERO_ITEMS"))