Megatest

Check-in [c8ec459e79]
Login
Overview
Comment:Reduce zombie cleanup, not clear it helps
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | v1.80-dbperformance
Files: files | file ages | folders
SHA1: c8ec459e7977f96daed577850f94d7beec3bc76d
User & Date: matt on 2023-02-03 21:51:17
Other Links: branch diff | manifest | tags
Context
2023-02-04
20:00
Change default number of db's to 10, couple minor fixes. Closed-Leaf check-in: 7b86032d25 user: matt tags: v1.80-dbperformance
2023-02-03
21:51
Reduce zombie cleanup, not clear it helps check-in: c8ec459e79 user: matt tags: v1.80-dbperformance
21:50
Cleanup droop curve a bit check-in: 07b36fe461 user: matt tags: v1.80-dbperformance
Changes

Modified launch.scm from [b249e00a3b] to [3243ef2278].

203
204
205
206
207
208
209
210

211
212
213
214
215
216
217
203
204
205
206
207
208
209

210
211
212
213
214
215
216
217







-
+







			(debug:print 0 *default-log-port* "WARNING: a prior step failed, stopping at " ezstep)))
                        (file-close status-file)
                        )

                        ))))))

(define (launch:monitor-job run-id test-id item-path fullrunscript ezsteps test-name tconfigreg exit-info m work-area runtlim misc-flags)
  (let* ((update-period (string->number (or (configf:lookup *configdat* "setup" "test-stats-update-period") "30")))
  (let* ((update-period (string->number (or (configf:lookup *configdat* "setup" "test-stats-update-period") "60")))
         (start-seconds (current-seconds))
	 (calc-minutes  (lambda ()
			  (inexact->exact 
			   (round 
			    (- 
			     (current-seconds) 
			     start-seconds)))))
257
258
259
260
261
262
263

264

265
266
267
268
269
270
271
257
258
259
260
261
262
263
264

265
266
267
268
269
270
271
272







+
-
+







         ((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)
        (if (common:low-noise-print 600 "run zombie") ;; every five minutes is plenty
        (launch:handle-zombie-tests run-id)
	    (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)))
	  )