Megatest

Check-in [143fb8134e]
Login
Overview
Comment:Merged earlier change
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | v1.80
Files: files | file ages | folders
SHA1: 143fb8134ee1266091c30ba2422bf49d0edc721c
User & Date: mrwellan on 2023-01-10 13:50:43
Other Links: branch diff | manifest | tags
Context
2023-01-11
07:14
Merged fork check-in: 43e394881f user: mrwellan tags: v1.80
2023-01-10
13:50
Merged earlier change check-in: 143fb8134e user: mrwellan tags: v1.80
2023-01-06
09:23
Synced mtargs with opensrc. Reduced noise in server output check-in: 7f6ea52b90 user: mrwellan tags: v1.80
2022-12-30
16:22
Changed version to 1.8004 check-in: e506ed709e user: mmgraham tags: v1.80, v1.8004
Changes

Modified db.scm from [3cf6da4c07] to [de020dc99f].

411
412
413
414
415
416
417
418

419
420
421
422
423













424
425
426
427

428
429
430
431
432
433
434
411
412
413
414
415
416
417

418
419




420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435

436
437
438
439
440
441
442
443







-
+

-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+



-
+







  (let* ((dbdat (db:open-db dbstruct #f db:initialize-main-db))
	 (data-synced       0) ;; count of changed records
    (tmp-area       (common:get-db-tmp-area))
    (dbfiles        (glob (conc tmp-area"/.megatest/*.db")))
    (sync-durations (make-hash-table))
    (no-sync-db        (db:open-no-sync-db)))
    (for-each
     (lambda (file)
     (lambda (file) ;; tmp db file
       (debug:print-info 3 *default-log-port* "file: " file)
       (let* ((fname       (conc (pathname-file file) ".db"))
	      (fulln       (conc *toppath*"/.megatest/"fname))
	      (time1       (if (file-exists? file)
			       (file-modification-time file)
       (let* ((fname       (conc (pathname-file file) ".db")) ;; fname is tmp db file
              (wal-file (conc fname "-wal"))
              (shm-file (conc fname "-shm"))
	      (fulln       (conc *toppath*"/.megatest/"fname)) ;; fulln is nfs db name
              (wal-time     (if (file-exists? wal-file)             
			       (file-modification-time wal-file)
                               0))
              (shm-time     (if (file-exists? shm-file)             
			       (file-modification-time shm-file)
                               0))

	      (time1       (if (file-exists? file)              ;; time1 is the max itime of the tmp db, -wal and -shm files.
			       (max (file-modification-time file) wal-time shm-time)
			       (begin
				 (debug:print-info 2 *default-log-port* "Sync - I do not see file "file)
			   1)))
	      (time2       (if (file-exists? fulln)
	      (time2       (if (file-exists? fulln)             ;; time2 is nfs file time
			       (file-modification-time fulln)
			       (begin
				 (debug:print-info 2 *default-log-port* "Sync - I do not see file "fulln)
				 0)))
	      (changed      (> (- time1 time2) (+ (random 5) 1)))  ;; it has been at some few seconds since last synced
	      (changed10    (> (- time1 time2) 10)) ;; it has been at least ten seconds since sync'd
	      (jfile-exists (file-exists? (conc file"-journal"))) ;; i.e. are we busy?
479
480
481
482
483
484
485
486
487

488
489
490
491
492
493
494
495
496
497
498
499
500
501
502



503
504
505
506
507
508
509
488
489
490
491
492
493
494


495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520







-
-
+















+
+
+







    (servers (server:choose-server *toppath* 'all-valid)) ;; (server:get-list *toppath*))
    (src-area (if old2new *toppath* tmp-area))
    (dest-area (if old2new tmp-area *toppath*))
    (dbfiles        (if old2new (glob (conc *toppath* "/.megatest/*.db")) (glob (conc tmp-area "/.megatest/*.db"))))
    (keys (db:get-keys dbstruct))
    (sync-durations (make-hash-table)))


    (if killservers
    (if (and killservers servers)
      (begin
       	  (for-each
	   (lambda (server)
             (handle-exceptions
             exn
             (begin 
               (debug:print-info 0 *default-log-port*  "Unable to get host and/or port from " server ", exn=" exn)     
               #f)
	     (match-let (((mod-time host port start-time server-id pid) server))
	       (if (and host pid)
		   (tasks:kill-server host pid)))))
	   servers)
          (delete-file* (common:get-sync-lock-filepath))
      )
    )

    (if (not dbfiles)
      (debug:print-error 0 *default-log-port* "no dbfiles found in " (conc *toppath* "/.megatest"))
    (for-each
     (lambda (srcfile)
       (debug:print-info 3 *default-log-port* "file: " srcfile)
       (let* ((fname (conc (pathname-file srcfile) ".db"))
              (basename (pathname-file srcfile))
              (run-id (if (string= basename "main") #f (string->number basename)))
	      (destfile (conc dest-area "/.megatest/" fname))
520
521
522
523
524
525
526

527
528
529
530
531
532
533
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545







+







			   (debug:print-info 0 *default-log-port* "Sync - I do not see file " destfile)
			   0)))
	      (changed ( < (- time2 time1) 6.0)) ;; dest db not updated within last 6 seconds

      (do-cp (cond
		      ((not (file-exists? destfile)) ;; shouldn't happen, but this might recover
		       (debug:print-info 2 *default-log-port* "File " destfile " not found. Copying "srcfile" to "destfile)
                       ;; TODO: Need to fix this for WAL mod. Can't just copy.
                       (system (conc "/bin/mkdir -p " dest-directory))
                       (system (conc "/bin/cp " srcfile " " destfile))
		       #t)
		      (changed ;; (and changed
		       ;; (> (- (current-seconds) time1) 3)) ;; if file is changed and three seconds have passed.
		       #t)
		      ((and changed *time-to-exit*) ;; last sync
557
558
559
560
561
562
563

564
565
566
567
568
569
570
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583







+







               )
	       (hash-table-set! sync-durations (conc srcfile ".db") (- (current-milliseconds) start-time)))
	     (debug:print-info 2 *default-log-port* "skipping delta sync. " srcfile " is up to date")
          )
       )
     )
     dbfiles
    )
    )
    data-synced
  )
)



2303
2304
2305
2306
2307
2308
2309
2310
2311


2312
2313
2314
2315
2316
2317
2318
2316
2317
2318
2319
2320
2321
2322


2323
2324
2325
2326
2327
2328
2329
2330
2331







-
-
+
+







    (db:with-db dbstruct run-id #f
		(lambda (dbdat db)
		  (sqlite3:for-each-row
		   (lambda (run-id testname item-path state status)
		     ;; id,run_id,testname,state,status,event_time,host,cpuload,diskfree,uname,rundir,item_path,run_duration,final_logf,comment
		     (set! res (vector test-id run-id testname state status -1 "" -1 -1 "" "-" item-path -1 "-" "-")))
		   db 
		   "SELECT run_id,testname,item_path,state,status FROM tests WHERE id=?;" 
		   test-id)))
		   "SELECT run_id,testname,item_path,state,status FROM tests WHERE id=? and run_id=?;" 
		   test-id run-id)))
    res))

;; get a useful subset of the tests data (used in dashboard
;; use db:mintest-get-{id ,run_id,testname ...}
;;
(define (db:get-tests-for-run-mindata dbstruct run-id testpatt states statuses not-in)
  (db:get-tests-for-run dbstruct run-id testpatt states statuses #f #f not-in #f #f "id,run_id,testname,state,status,event_time,item_path" 0 #f))

Modified dbfile.scm from [789cb4b125] to [fc2ee87f83].

334
335
336
337
338
339
340
341

342
343
344
345
346
347
348
349
350
351
352
353
354
355
356

357
358
359
360
361
362
363
334
335
336
337
338
339
340

341
342
343
344
345
346
347
348
349
350
351
352
353
354
355

356
357
358
359
360
361
362
363







-
+














-
+







	 (delay-time (* (- 51 tries-left) 1.1))
      	 (write-access (file-write-access? fname))
         (dir-access (file-write-access? (pathname-directory fname)))
         (retry      (lambda ()
		       (thread-sleep! delay-time)
		       (if (> tries-left 0)
			   (dbfile:cautious-open-database fname init-proc
							  sync-mode: sync-mode journal-mode: journal-mode
							  sync-mode: sync-mode journal-mode
							  (- tries-left 1))))))
    (assert (>= tries-left 0) (conc "FATAL: too many attempts in dbfile:cautious-open-database of "fname", giving up."))
    
    (if (and (file-write-access? fname)
	     (file-exists? busy-file))
	(begin
	  (if (common:low-noise-print 120 busy-file)
	      (dbfile:print-err "INFO: dbfile:cautious-open-database: journal file "
				busy-file" exists, trying again in few seconds."))
	  (thread-sleep! 1)
	  (if (eq? tries-left 2)
	      (begin
	  	(dbfile:print-err "INFO: forcing journal rollup "busy-file)
	  	(dbfile:brute-force-salvage-db fname)))
	  (dbfile:cautious-open-database fname init-proc sync-mode: sync-mode journal-mode: journal-mode (- tries-left 1)))
	  (dbfile:cautious-open-database fname init-proc sync-mode journal-mode (- tries-left 1)))
	
	(let* ((result (condition-case
		        (if dir-access
			    (dbfile:with-simple-file-lock
			     (conc fname ".lock")
			     (lambda ()
			       (let* ((db-exists (file-exists? fname))

Modified http-transport.scm from [d351abf6cf] to [e5186c35f3].

151
152
153
154
155
156
157
158

159
160
161
162
163
164
165
151
152
153
154
155
156
157

158
159
160
161
162
163
164
165







-
+







	(config-use-proxy (equal? (configf:lookup *configdat* "client" "use-http_proxy") "yes")))
    (if (not config-use-proxy)
	(determine-proxy (constantly #f)))
    (debug:print-info 0 *default-log-port* "http-transport:try-start-server time=" (seconds->time-string (current-seconds)) " ipaddrsstr=" ipaddrstr " portnum=" portnum " config-hostname=" config-hostname)
    (handle-exceptions
	exn
	(begin
	  (print-error-message exn)
	  ;; (print-error-message exn)
	  (if (< portnum 64000)
	      (begin 
		(debug:print 0 *default-log-port* "WARNING: attempt to start server failed. Trying again ...")
		(debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
		(debug:print 5 *default-log-port* "exn=" (condition->list exn))
		(portlogger:open-run-close portlogger:set-failed portnum)
		(debug:print 0 *default-log-port* "WARNING: failed to start on portnum: " portnum ", trying next port")
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463

464

465
466
467

468
469
470
471
472

473
474
475
476


477
478
479
480
481
482
483
446
447
448
449
450
451
452






453
454
455
456
457
458

459
460
461

462





463




464
465
466
467
468
469
470
471
472







-
-
-
-
-
-





+
-
+


-
+
-
-
-
-
-
+
-
-
-
-
+
+







				      (print "started: "(seconds->year-week/day-time (current-seconds))))))
				(set! *on-exit-procs* (cons
						       (lambda ()
							 (delete-file* servinf))
						       *on-exit-procs*))
				;; put data about this server into a simple flat file host.port
				(debug:print-info 0 *default-log-port* "Received server alive signature")
                                #;(common:save-pkt `((action . alive)
                                                   (T      . server)
                                                   (pid    . ,(current-process-id))
                                                   (ipaddr . ,(car sdat))
                                                   (port   . ,(cadr sdat)))
                                                 *configdat* #t)
				sdat)
                              (begin
				(debug:print-info 0 *default-log-port* "Still waiting, last-sdat=" last-sdat)
                                (sleep 4)
				(if (> (- (current-seconds) start-time) 120) ;; been waiting for two minutes
                                    (if sdat 
				    (let* ((ipaddr  (car sdat))
				      (let* ((ipaddr  (car sdat))
					   (port    (cadr sdat))
					   (servinf (conc (server:get-servinfo-dir *toppath*)"/"ipaddr":"port)))
				      (debug:print-error 0 *default-log-port* "transport appears to have died, exiting server")
				        (debug:print-error 0 *default-log-port* "transport appears to have died, exiting server")
				      ;; (delete-file* servinf) ;; handled by on-exit, can be removed
                                      #;(common:save-pkt `((action . died)
                                                         (T      . server)
                                                         (pid    . ,(current-process-id))
                                                         (ipaddr . ,(car sdat))
				      (exit))
                                                         (port   . ,(cadr sdat))
                                                         (msg    . "Transport died?"))
						       *configdat* #t)
				      (exit))
                                      (exit)
                                    )
				    (loop start-time
					  (equal? sdat last-sdat)
					  sdat)))))))
	 (iface       (car server-info))
         (port        (cadr server-info))
         (last-access 0)
	 (server-timeout (server:expiration-timeout))
574
575
576
577
578
579
580
581

582
583
584

585
586
587
588

589
590
591
592
593
594
595
563
564
565
566
567
568
569

570
571
572

573
574
575
576

577
578
579
580
581
582
583
584







-
+


-
+



-
+







		(handle-exceptions
		    exn
		    (debug:print 0 *default-log-port* "ERROR: Failed to change timestamp on info file " servinfofile ". Are you out of space on that disk? exn=" exn)
		    (if (and ;; (< (- (current-seconds) server-start-time) 600) ;; run for ten minutes for experiment, 3600 thereafter
			     (not *server-overloaded*)
			     (file-exists? servinfofile))
			(change-file-times servinfofile curr-time curr-time)))
		(if (or (common:low-noise-print 120 "start new server")
		(if (and (common:low-noise-print 120 "start new server")
			(> *api-process-request-count* 50)) ;; if this server is kind of busy start up another
		    (begin
		      (debug:print-info 0 *default-log-port* "Server is busy, parallel-api-count "*api-process-request-count*", start another if possible...")
		      (debug:print-info 0 *default-log-port* "Server is busy, api-count "*api-process-request-count*", start another if possible...")
		      (server:kind-run *toppath*)
		      (if (> *api-process-request-count* 100)
			  (begin
			    (debug:print-info 0 *default-log-port* "Server is overloaded at parallel-api-count="*api-process-request-count*", removing "servinfofile) 
			    (debug:print-info 0 *default-log-port* "Server is overloaded at api-count=" *api-process-request-count*", removing "servinfofile) 
			    (delete-file* servinfofile)))))))
          (loop 0 server-state bad-sync-count (current-milliseconds)))
         (else
          (debug:print-info 0 *default-log-port* "Server timed out. seconds since last db access: " (- (current-seconds) last-access))
          (http-transport:server-shutdown port)))))))

(define (http-transport:server-shutdown port)

Modified megatest-version.scm from [c1e20d2fad] to [1632a50df8].

16
17
18
19
20
21
22
23

16
17
18
19
20
21
22

23







-
+
;;     along with Megatest.  If not, see <http://www.gnu.org/licenses/>.

;; Always use two or four digit decimal
;; 1.01, 1.02...1.10,1.11,1.1101 ... 1.99,2.00..

;; (declare (unit megatest-version))

(define megatest-version 1.8003)
(define megatest-version 1.8004)

Modified megatest.scm from [81aaa85fa0] to [79d9696058].

652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
652
653
654
655
656
657
658













659
660
661
662
663
664
665







-
-
-
-
-
-
-
-
-
-
-
-
-







		      (original-exit exit-code)))))

;; for some switches always print the command to stderr
;;
(if (args:any? "-run" "-runall" "-remove-runs" "-set-state-status" "-kill-runs" "-kill-rerun")
    (debug:print 0 *default-log-port* (string-intersperse (argv) " ")))

;; some switches imply homehost. Exit here if not on homehost
;;
(let ((homehost-required  (list "-cleanup-db")))
  (if (apply args:any? homehost-required)
      (if (not (server:choose-server *toppath* 'home?))
	  (for-each
	   (lambda (switch)
	     (if (args:get-arg switch)
		 (begin
		   (debug:print 0 *default-log-port* "ERROR: you must be on the homehost to run with " switch
				", you can move homehost by removing the .homehost file but this will disrupt any runs in progress.")
		   (exit 1))))
	   homehost-required))))

;;======================================================================
;; Misc setup stuff
;;======================================================================

(debug:setup)

2319
2320
2321
2322
2323
2324
2325




2326
2327
2328
2329
2330
2331
2332
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323







+
+
+
+








(if (args:get-arg "-cleanup-db")
    (begin
      (if (not (launch:setup))
	  (begin
	    (debug:print 0 *default-log-port* "Failed to setup, exiting") 
	    (exit 1)))
      (if (not (server:choose-server *toppath* 'home?))
	  (begin
	    (debug:print 0 *default-log-port* "Servers are not running on this host or no servers alive. Cannot run cleanup-db")
	    (exit 1)))
      (let ((dbstructs (db:setup #f)))
        (common:cleanup-db dbstructs))
      (set! *didsomething* #t)))

(if (args:get-arg "-mark-incompletes")
    (begin
      (if (not (launch:setup))

Modified portlogger.scm from [36a4964f50] to [8344cdf37f].

62
63
64
65
66
67
68
69

70
71
72
73
74
75
76
62
63
64
65
66
67
68

69
70
71
72
73
74
75
76







-
+







	 (avail  (tasks:wait-on-journal fname 10))) ;; wait up to about 10 seconds for the journal to go away
    (handle-exceptions
     exn
     (begin
       ;; (release-dot-lock fname)
       (debug:print-error 0 *default-log-port* "portlogger:open-run-close failed. " proc " " params)
       (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
       (debug:print 5 *default-log-port* "exn=" (condition->list exn))
       ;; (debug:print 5 *default-log-port* "exn=" (condition->list exn))
       (if (common:file-exists? fname)(delete-file fname)) ;; brutally get rid of it
       (print-call-chain (current-error-port)))
     (let* (;; (lock   (obtain-dot-lock fname 2 9 10))
	    (db     (portlogger:open-db fname))
	    (res    (apply proc db params)))
       (sqlite3:finalize! db)
       ;; (release-dot-lock fname)

Modified server.scm from [df559aacc0] to [1caa4a85a3].

401
402
403
404
405
406
407
408

409
410
411
412
413
414
415
401
402
403
404
405
406
407

408
409
410
411
412
413
414
415







-
+







		 (thread-sleep! ( + 1 idletime))
		 (server:wait-for-server-start-last-flag areapath)))))))

;; oldest server alive determines host then choose random of youngest
;; five servers on that host
;;
(define (server:get-servers-info areapath)
  (assert *toppath* "FATAL: server:get-servers-info called before *toppath* has been set.")
  ;; (assert *toppath* "FATAL: server:get-servers-info called before *toppath* has been set.")
  (let* ((servinfodir (server:get-servinfo-dir areapath))) ;; (conc *toppath*"/.servinfo")))
    (if (not (file-exists? servinfodir))
	(create-directory servinfodir))
    (let* ((allfiles    (glob (conc servinfodir"/*")))
	   (res         (make-hash-table)))
      (for-each
       (lambda (f)
447
448
449
450
451
452
453




454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470











471
472
473
474
475
476
477
478


479
480
481
482
483
484
485
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502







+
+
+
+

















+
+
+
+
+
+
+
+
+
+
+








+
+







     ((and havehome wearehome (< (length thisservrs) 20)) #t) ;; we are home and less than 20 servers, ok to start another
     ((and havehome (not wearehome)) #f)     ;; we are not the home host
     ((and havehome wearehome (>= (length thisservrs) 20)) #f) ;; have enough running
     (else
      (debug:print 0 *default-log-port* "WARNING: Unrecognised scenario, servrs="servrs", thishostip="thishostip", thisservrs="thisservrs)
      #t))))
	 

(define server-last-start 0)


;; oldest server alive determines host then choose random of youngest
;; five servers on that host
;;
;; mode:
;;   best - get best server (random of newest five)
;;   home - get home host based on oldest server
;;   info - print info
(define (server:choose-server areapath #!optional (mode 'best))
  ;; age is current-starttime
  ;; find oldest alive
  ;;   1. sort by age ascending and ping until good
  ;; find alive rand from youngest
  ;;   1. sort by age descending
  ;;   2. take five
  ;;   3. check alive, discard if not and repeat
  ;; first we clean up old server files
  (server:clean-up-old areapath)
  (let* ((since-last (- (current-seconds) server-last-start))
        (server-start-delay 10))     
    (if ( < (- (current-seconds) server-last-start) 10 )
      (begin
        (debug:print 2 *default-log-port* "server:choose-server: seconds since last server start: " (- (current-seconds) server-last-start))
        (debug:print 2 *default-log-port* "server:choose-server: last server start less than " server-start-delay " seconds ago. Sleeping " server-start-delay " seconds")
        (thread-sleep! server-start-delay)
      )
      (debug:print 2 *default-log-port* "server:choose-server: seconds since last server start: " (- (current-seconds) server-last-start))
    )
  )
  (let* ((serversdat  (server:get-servers-info areapath))
	 (servkeys    (hash-table-keys serversdat))
	 (by-time-asc (if (not (null? servkeys)) ;; NOTE: Oldest is last
			  (sort servkeys ;; list of "host:port"
				(lambda (a b)
				  (>= (list-ref (hash-table-ref serversdat a) 2)
				      (list-ref (hash-table-ref serversdat b) 2))))
			  '())))
    (debug:print 2 *default-log-port* "server:choose-server: serversdat: " serversdat)
    (debug:print 2 *default-log-port* "server:choose-server: servkeys: " servkeys)
    (if (not (null? by-time-asc))
	(let* ((oldest     (last by-time-asc))
	       (oldest-dat (hash-table-ref serversdat oldest))
	       (host       (list-ref oldest-dat 0))
	       (all-valid  (filter (lambda (x)
				     (equal? host (list-ref (hash-table-ref serversdat x) 0)))
				   by-time-asc))
512
513
514
515
516
517
518

519

520
521
522
523
524
525
526
529
530
531
532
533
534
535
536

537
538
539
540
541
542
543
544







+
-
+







			  (hash-table-ref serversdat (list-ref best-ten (random len)))))
	    ((count)(length all-valid))
	    (else
	     (debug:print 0 *default-log-port* "ERROR: invalid command "mode)
	     #f)))
	(begin
	  (server:run areapath)
          (set! server-last-start (current-seconds))
	  (thread-sleep! 3)
	  ;; (thread-sleep! 3)
	  (case mode
	    ((homehost) (cons #f #f))
	    (else	#f))))))

(define (server:get-servinfo-dir areapath)
  (let* ((spath (conc areapath"/.servinfo")))
    (if (not (file-exists? spath))
585
586
587
588
589
590
591

592

593
594
595
596
597
598
599
603
604
605
606
607
608
609
610

611
612
613
614
615
616
617
618







+
-
+







(define (server:start-and-wait areapath #!key (timeout 60))
  (let ((give-up-time (+ (current-seconds) timeout)))
    (let loop ((server-info (server:check-if-running areapath))
	       (try-num    0))
      (if (or server-info
	      (> (current-seconds) give-up-time)) ;; server-url will be #f if no server available.
	  (server:record->url server-info)
	  (let* ( (servers (server:choose-server areapath 'all-valid))
	  (let ((num-ok (length (server:choose-server areapath 'all-valid))))
                (num-ok (if servers (length (server:choose-server areapath 'all-valid)) 0)))
	    (if (and (> try-num 0)  ;; first time through simply wait a little while then try again
		     (< num-ok 1))  ;; if there are no decent candidates for servers then try starting a new one
		(server:run areapath))
	    (thread-sleep! 5)
	    (loop (server:check-if-running areapath)
		  (+ try-num 1)))))))