345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
|
;; (car processes))
;; (system (conc "kill -9 -" pid))))
(begin
(debug:print 0 "WARNING: Request received to kill job but problem with process, attempting to kill manager process")
(tests:test-set-status! test-id "KILLED" "FAIL"
(args:get-arg "-m") #f)
(sqlite3:finalize! tdb)
(exit 1))))
(set! kill-tries (+ 1 kill-tries))
(mutex-unlock! m)))
;; (sqlite3:finalize! db)
(if keep-going
(begin
(thread-sleep! 3) ;; (+ 3 (random 6))) ;; add some jitter to the call home time to spread out the db accesses
(if keep-going
(loop (calc-minutes)))))))))) ;; NOTE: Checking twice for keep-going is intentional
(th1 (make-thread monitorjob "monitor job"))
(th2 (make-thread runit "run job")))
(set! job-thread th2)
(thread-start! th1)
(thread-start! th2)
(thread-join! th2)
(set! keep-going #f)
(thread-join! th1)
;; (thread-sleep! 1)
;; (thread-terminate! th1) ;; Not sure if this is a good idea
(thread-sleep! 1) ;; give thread th1 a chance to be done TODO: Verify this is needed. At 0.1 I was getting fail to stop, increased to total of 1.1 sec.
(mutex-lock! m)
(let* ((item-path (item-list->path itemdat))
(testinfo (cdb:get-test-info-by-id *runremote* test-id))) ;; )) ;; run-id test-name item-path)))
;; Am I completed?
(if (member (db:test-get-state testinfo) '("REMOTEHOSTSTART" "RUNNING")) ;; NOTE: It should *not* be REMOTEHOSTSTART but for reasons I don't yet understand it sometimes gets stuck in that state ;; (not (equal? (db:test-get-state testinfo) "COMPLETED"))
(let ((new-state (if kill-job? "KILLED" "COMPLETED") ;; (if (eq? (vector-ref exit-info 2) 0) ;; exited with "good" status
;; "COMPLETED"
|
|
>
|
>
>
|
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
|
;; (car processes))
;; (system (conc "kill -9 -" pid))))
(begin
(debug:print 0 "WARNING: Request received to kill job but problem with process, attempting to kill manager process")
(tests:test-set-status! test-id "KILLED" "FAIL"
(args:get-arg "-m") #f)
(sqlite3:finalize! tdb)
(exit 1) ;; IS THIS NECESSARY OR WISE???
)))
(set! kill-tries (+ 1 kill-tries))
(mutex-unlock! m)))
;; (sqlite3:finalize! db)
(if keep-going
(begin
(thread-sleep! 3) ;; (+ 3 (random 6))) ;; add some jitter to the call home time to spread out the db accesses
(if keep-going
(loop (calc-minutes)))))))
(tests:update-central-meta-info test-id (get-cpu-load) (get-df (current-directory))(calc-minutes) #f #f)))) ;; NOTE: Checking twice for keep-going is intentional
(th1 (make-thread monitorjob "monitor job"))
(th2 (make-thread runit "run job")))
(set! job-thread th2)
(thread-start! th1)
(thread-start! th2)
(thread-join! th2)
(set! keep-going #f)
(thread-join! th1)
;; (thread-sleep! 1)
;; (thread-terminate! th1) ;; Not sure if this is a good idea
(thread-sleep! 1) ;; give thread th1 a chance to be done TODO: Verify this is needed. At 0.1 I was getting fail to stop, increased to total of 1.1 sec.
;; (tests:update-central-meta-info test-id cpuload diskfree minutes #f #f)
(mutex-lock! m)
(let* ((item-path (item-list->path itemdat))
(testinfo (cdb:get-test-info-by-id *runremote* test-id))) ;; )) ;; run-id test-name item-path)))
;; Am I completed?
(if (member (db:test-get-state testinfo) '("REMOTEHOSTSTART" "RUNNING")) ;; NOTE: It should *not* be REMOTEHOSTSTART but for reasons I don't yet understand it sometimes gets stuck in that state ;; (not (equal? (db:test-get-state testinfo) "COMPLETED"))
(let ((new-state (if kill-job? "KILLED" "COMPLETED") ;; (if (eq? (vector-ref exit-info 2) 0) ;; exited with "good" status
;; "COMPLETED"
|