Megatest

# make install CSCOPTS='-accumulate-profile -profile-name $(PWD)/profile-ww$(shell date +%V.%u)'
# rm <files>.o ; make install CSCOPTS='-profile' ; ... ;  chicken-profile | less

PREFIX=$(PWD)
CSCOPTS=
INSTALL=install
SRCFILES = common.scm items.scm launch.scm \
   ods.scm runconfig.scm server.scm configf.scm \
   db.scm keys.scm margs.scm megatest-version.scm \
   process.scm runs.scm tasks.scm tests.scm genexample.scm \
   http-transport.scm filedb.scm tdb.scm \
   client.scm synchash.scm daemon.scm mt.scm \
   client.scm daemon.scm mt.scm \
   ezsteps.scm lock-queue.scm sdb.scm \
   rmt.scm api.scm \
   portlogger.scm archive.scm env.scm diff-report.scm cgisetup/models/pgdb.scm

# Eggs to install (straightforward ones)
EGGS=matchable readline apropos base64 regex-literals format regex-case test coops trace csv \
dot-locking posix-utils posix-extras directory-utils hostinfo tcp-server rpc csv-xml fmt \

;;======================================================================
;; Copyright 2006-2012, Matthew Welland.
;; 
;;  This program is made available under the GNU GPL version 2.0 or
;;  greater. See the accompanying file COPYING for details.
;; 
;;  This program is distributed WITHOUT ANY WARRANTY; without even the
;;  implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
;;  PURPOSE.
;;======================================================================

(use srfi-1 posix regex-case base64 format dot-locking csv-xml z3 sql-de-lite hostinfo md5 message-digest typed-records directory-utils stack
     matchable regex posix srfi-18 extras)
     matchable regex posix srfi-18 extras
     pkts (prefix dbi dbi:))

(import (prefix sqlite3 sqlite3:))
(import (prefix base64 base64:))

(declare (unit common))

(include "common_records.scm")

(define (common:get-last-run-version-number)
  (string->number 
   (substring (common:get-last-run-version) 0 6)))

(define (common:set-last-run-version)
  (rmt:set-var "MEGATEST_VERSION" (common:version-signature)))

;; postive number if megatest version > db version
;; negative number if megatest version < db version
(define (common:version-db-delta)
         (- megatest-version (common:get-last-run-version-number)))

(define (common:version-changed?)
  (not (equal? (common:get-last-run-version)
	       (common:version-signature))))
               (common:version-signature))))

(define (common:api-changed?)
  (not (equal? (substring (->string megatest-version) 0 4)
               (substring (common:get-last-run-version) 0 4))))
  
;; Move me elsewhere ...
;; RADT => Why do we meed the version check here, this is called only if version misma
;;
(define (common:cleanup-db dbstruct)
  (db:multi-db-sync 
   dbstruct
   'schema
   ;; 'new2old
   'killservers
   'dejunk
   'adj-target
   ;; 'old2new
   'new2old
   )
  (if (common:version-changed?)
  (if (common:api-changed?)
      (common:set-last-run-version)))

;; Rotate logs, logic: 
;;                 if > 500k and older than 1 week:
;;                     remove previous compressed log and compress this log
;; WARNING: This proc operates assuming that it is in the directory above the
;;          logs directory you wish to log-rotate.

   "logs"))

;; Force a megatest cleanup-db if version is changed and skip-version-check not specified
;; Do NOT check if not on homehost!
;;
(define (common:exit-on-version-changed)
  (if (common:on-homehost?)
      (if (common:version-changed?)
      (if (common:api-changed?)
	  (let* ((mtconf (conc (get-environment-variable "MT_RUN_AREA_HOME") "/megatest.config"))
                (dbfile (conc (get-environment-variable "MT_RUN_AREA_HOME") "/megatest.db"))
                (read-only (not (file-write-access? dbfile)))
                (dbstruct (db:setup #t)))
	    (debug:print 0 *default-log-port*
			 "WARNING: Version mismatch!\n"
			 "   expected: " (common:version-signature) "\n"

      #f ;; I don't really care why this failed (at least for now)
    (delete-file* fname)))

;;======================================================================
;; S T A T E S   A N D   S T A T U S E S
;;======================================================================

;; BBnote: *common:std-states* - dashboard filter control and test control state buttons defined here; used in set-fields-panel and dboard:make-controls
(define *common:std-states*   
(define *common:std-states*   ;; for toggle buttons in dashboard
  '((0 "ARCHIVED")
    (1 "STUCK")
    (2 "KILLREQ")
    (3 "KILLED")
    (4 "NOT_STARTED")
    (5 "COMPLETED")
    (6 "LAUNCHED")
    (7 "REMOTEHOSTSTART")
    (8 "RUNNING")
    ))

;; BBnote: *common:std-statuses* dashboard filter control and test control status buttons defined here; used in set-fields-panel and dboard:make-controls
(define *common:std-statuses*
  '(;; (0 "DELETED")
    (1 "n/a")
    (2 "PASS")
    (3 "SKIP")
    (4 "WARN")
    (5 "WAIVED")
    (6 "CHECK")
    (7 "STUCK/DEAD")
    (8 "FAIL")
    (9 "ABORT")))

(define *common:ended-states*       ;; states which indicate the test is stopped and will not proceed
  '("COMPLETED" "ARCHIVED" "KILLED" "KILLREQ" "STUCK" "INCOMPLETE"))

(define *common:badly-ended-states* ;; these roll up as CHECK, i.e. results need to be checked
  '("KILLED" "KILLREQ" "STUCK" "INCOMPLETE" "DEAD"))

;; BBnote: *common:running-states* used from db:set-state-status-and-roll-up-items
(define *common:running-states*     ;; test is either running or can be run
  '("RUNNING" "REMOTEHOSTSTART" "LAUNCHED"))
  '("RUNNING" "REMOTEHOSTSTART" "LAUNCHED" "STARTED"))

(define *common:cant-run-states*    ;; These are stopping conditions that prevent a test from being run
  '("COMPLETED" "KILLED" "UNKNOWN" "INCOMPLETE" "ARCHIVED"))

(define *common:not-started-ok-statuses* ;; if not one of these statuses when in not_started state treat as dead
  '("n/a" "na" "PASS" "FAIL" "WARN" "CHECK" "WAIVED" "DEAD" "SKIP"))

      (getenv "MT_TESTSUITE_NAME")
      (if (string? *toppath* )
          (pathname-file *toppath*)
          #f))) ;; (pathname-file (current-directory)))))

(define common:get-area-name common:get-testsuite-name)

(define (common:get-db-tmp-area)
(define (common:get-db-tmp-area . junk)
  (if *db-cache-path*
      *db-cache-path*
      (if *toppath*
	  (let ((dbpath (create-directory (conc "/tmp/" (current-user-name)
						"/megatest_localdb/"
						(common:get-testsuite-name) "/"
						(string-translate *toppath* "/" ".")) #t)))
	    (set! *db-cache-path* dbpath)
	    dbpath)
      (if *toppath* ;; common:get-create-writeable-dir
	  (handle-exceptions
	      exn
	      (begin
		(debug:print-error 0 *default-log-port* "Couldn't create path to " dbdir)
		(exit 1))
	    (let ((dbpath (common:get-create-writeable-dir
			   (list (conc "/tmp/" (current-user-name)
				       "/megatest_localdb/"
				       (common:get-testsuite-name) "/"
				       (string-translate *toppath* "/" ".")))))) ;;  #t))))
	      (set! *db-cache-path* dbpath)
	      dbpath))
	  #f)))

(define (common:get-area-path-signature)
  (message-digest-string (md5-primitive) *toppath*))

(define (common:get-signature str)
  (message-digest-string (md5-primitive) str))

	 (mthome-cfgfile (conc *toppath* "/.mtviews.config")))
    (if (file-exists? mthome-cfgfile)
	(read-config mthome-cfgfile view-cfgdat #t))
    ;; we load the home dir file AFTER the MTRAH file so the user can clobber settings when running the dashboard in read-only areas
    (if (file-exists? home-cfgfile)
	(read-config home-cfgfile view-cfgdat #t))
    view-cfgdat))

;;======================================================================
;; Manage pkts, used in servers, tests and likely other contexts so put
;; in common
;;======================================================================

(define common:pkt-spec
  '((server . ((action    . a)
	       (pid       . d)
	       (ipaddr    . i)
	       (port      . p)))
    			  
    (test   . ((cpuuse    . c)
	       (diskuse   . d)
	       (item-path . i)
	       (runname   . r)
	       (state     . s)
	       (target    . t)
	       (status    . u)))))

(define (common:get-pkts-dirs mtconf use-lt)
  (let* ((pktsdirs-str (or (configf:lookup mtconf "setup"  "pktsdirs")
			   (and use-lt
				(conc *toppath* "/lt/.pkts"))))
	 (pktsdirs  (if pktsdirs-str
			(string-split pktsdirs-str " ")
			#f)))
    pktsdirs))

(define (common:with-queue-db mtconf proc #!key (use-lt #f)(toppath-in #f))
  (let* ((pktsdirs (common:get-pkts-dirs mtconf use-lt))
	 (pktsdir  (if pktsdirs (car pktsdirs) #f))
	 (toppath  (or (configf:lookup mtconf "scratchdat" "toppath")
		       toppath-in))
	 (pdbpath  (or (configf:lookup mtconf "setup"  "pdbpath") pktsdir)))
    (if (not (and  pktsdir toppath pdbpath))
	(begin
	  (print "ERROR: settings are missing in your megatest.config for area management.")
	  (print "  you need to have pktsdir in the [setup] section."))
	(let* ((pdb  (open-queue-db pdbpath "pkts.db"
				    schema: '("CREATE TABLE groups (id INTEGER PRIMARY KEY,groupname TEXT, CONSTRAINT group_constraint UNIQUE (groupname));"))))
	  (proc pktsdirs pktsdir pdb)
	  (dbi:close pdb)))))

(define (common:load-pkts-to-db mtconf)
  (common:with-queue-db
   mtconf
   (lambda (pktsdirs pktsdir pdb)
     (for-each
      (lambda (pktsdir) ;; look at all
	(if (and (file-exists? pktsdir)
		 (directory? pktsdir)
		 (file-read-access? pktsdir))
	    (let ((pkts (glob (conc pktsdir "/*.pkt"))))
	      (for-each
	       (lambda (pkt)
		 (let* ((uuid    (cadr (string-match ".*/([0-9a-f]+).pkt" pkt)))
			(exists  (lookup-by-uuid pdb uuid #f)))
		   (if (not exists)
		       (let* ((pktdat (string-intersperse
				       (with-input-from-file pkt read-lines)
				       "\n"))
			      (apkt   (pkt->alist pktdat))
			      (ptype  (alist-ref 'T apkt)))
			 (add-to-queue pdb pktdat uuid (or ptype 'cmd) #f 0)
			 (debug:print 4 *default-log-port* "Added " uuid " of type " ptype " to queue"))
		       (debug:print 4 *default-log-port* "pkt: " uuid " exists, skipping...")
		       )))
	       pkts))))
      pktsdirs))))

(define (common:get-pkt-alists pkts)
  (map (lambda (x)
	 (alist-ref 'apkt x)) ;; 'pkta pulls out the alist from the read pkt
       pkts))

;; given list of pkts (alist mode) return list of D cards as Unix epoch, sorted descending
;; also delete duplicates by target i.e. (car pkt)
;;
(define (common:get-pkt-times pkts)
  (delete-duplicates
   (sort 
    (map (lambda (x)
	   `(,(alist-ref 't x) . ,(string->number (alist-ref 'D x))))
	 pkts)
    (lambda (a b)(> (cdr a)(cdr b))))      ;; sort descending
   (lambda (a b)(equal? (car a)(car b))))) ;; remove duplicates by target

	    (let ((match (assoc var sectdat)))
	      (if match ;; (and match (list? match)(> (length match) 1))
		  (cadr match)
		  #f))
	    ))
      #f))

;; use to have definitive setting:
;;  [foo]
;;  var yes
;;
;;  (configf:var-is? cfgdat "foo" "var" "yes") => #t
;;
(define (configf:var-is? cfgdat section var expected-val)
  (equal? (configf:lookup cfgdat section var) expected-val))

(define configf:lookup config-lookup)
(define configf:read-file read-config)

;; safely look up a value that is expected to be a number, return
;; a default (#f unless provided)
;;
(define (configf:lookup-number cfdat section varname #!key (default #f))
  (let* ((val (configf:lookup *configdat* section varname))
         (res (if val
                  (string->number (string-substitute "\\s+" "" val #t))
                  #f)))
    (cond
     (res  res)
     (val  (debug:print 0 *default-log-port* "ERROR: no number found for [" section "], " varname ", got: " val))
     (else default))))

(define (configf:section-vars cfgdat section)
  (let ((sectdat (hash-table-ref/default cfgdat section '())))
    (if (null? sectdat)
	'()
	(map car sectdat))))

;;     (filedb:get-path db id)))

;; NB// #f => return dbdir only
;;      (was planned to be;  zeroth db with name=main.db)
;; 
;; If run-id is #f return to create and retrieve the path where the db will live.
;;
(define (db:dbfile-path . junk) ;;  run-id)
  (let* ((dbdir           (common:get-db-tmp-area)))
(define db:dbfile-path common:get-db-tmp-area)
    (handle-exceptions
     exn
     (begin
       (debug:print-error 0 *default-log-port* "Couldn't create path to " dbdir)
       (exit 1))
     (if (not (directory? dbdir))(create-directory dbdir #t)))
    dbdir))

(define (db:set-sync db)
  (let ((syncprag (configf:lookup *configdat* "setup" "sychronous")))
    (sqlite3:execute db (conc "PRAGMA synchronous = " (or syncprag 0) ";")))) 

;; open an sql database inside a file lock
;; returns: db existed-prior-to-opening

(define (db:sync-tables tbls last-update fromdb todb . slave-dbs)
  (handle-exceptions
   exn
   (begin
     (debug:print 0 *default-log-port* "EXCEPTION: database probably overloaded or unreadable in db:sync-tables.")
     (print-call-chain (current-error-port))
     (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
     (print "exn=" (condition->list exn))
     (debug:print 5 *default-log-port* "exn=" (condition->list exn))
     (debug:print 0 *default-log-port* " status:  " ((condition-property-accessor 'sqlite3 'status) exn))
     (debug:print 0 *default-log-port* " src db:  " (db:dbdat-get-path fromdb))
     (for-each (lambda (dbdat)
		 (let ((dbpath (db:dbdat-get-path dbdat)))
		   (debug:print 0 *default-log-port* " dbpath:  " dbpath)
		   (if (not (db:repair-db dbdat))
		       (begin

	 (err-status ((condition-property-accessor 'sqlite3 'status #f) exn)))
     (case err-status
       ((busy)
	(thread-sleep! sleep-time))
       (else
	(debug:print 0 *default-log-port* "EXCEPTION: database probably overloaded or unreadable.")
	(debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	(print "exn=" (condition->list exn))
	(debug:print 5 *default-log-port* "exn=" (condition->list exn))
	(debug:print 0 *default-log-port* " status:  " ((condition-property-accessor 'sqlite3 'status) exn))
	(print-call-chain (current-error-port))
	(thread-sleep! sleep-time)
	(debug:print-info 0 *default-log-port* "trying db call one more time....this may never recover, if necessary kill process " (current-process-id) " on host " (get-host-name) " to clean up")))
     (apply open-run-close-exception-handling proc idb params))
   (apply open-run-close-no-exception-handling proc idb params)))

;; 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)
  ;; 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
  
  (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))

                                                                                      *common:not-started-ok-statuses*))))
								  state-status-counts)))
                            ;; (non-completes        (filter (lambda (x)
                            ;;                                 (not (equal? (dbr:counts-state x) "COMPLETED")))
                            ;;                               state-status-counts))
                            (all-curr-states      (common:special-sort  ;; worst -> best (sort of)
                                                   (delete-duplicates
                                                    (if (not (equal? state "DELETED"))
                                                    (cons state (map dbr:counts-state state-status-counts)))
                                                        (cons state (map dbr:counts-state state-status-counts))
                                                        (map dbr:counts-state state-status-counts)))
                                                   *common:std-states* >))
                            (all-curr-statuses    (common:special-sort  ;; worst -> best
                                                   (delete-duplicates
                                                    (if (not (equal? state "DELETED"))
                                                    (cons status (map dbr:counts-status state-status-counts)))
                                                        (cons status (map dbr:counts-status state-status-counts))
                                                        (map dbr:counts-status state-status-counts)))
                                                   *common:std-statuses* >))
			    (non-completes     (filter (lambda (x)
							 (not (equal? x "COMPLETED")))
						       all-curr-states))
			    (num-non-completes (length non-completes))
                            
                            (newstate          (cond
						((> running 0)
						 "RUNNING") ;; anything running, call the situation running
						((> bad-not-started 0)  ;; we have an ugly situation, it is completed in the sense we cannot do more.
						 "COMPLETED") 
						((> num-non-completes 0) ;;
						 (car non-completes))  ;;  (remove (lambda (x)(equal? "COMPLETED" x)) all-curr-states)))
                                                ;; only rollup DELETED if all DELETED
						(else
						 (car all-curr-states))))
			                       ;; (if (> running 0)
                                               ;;     "RUNNING"
                                               ;;     (if (> bad-not-started 0)
                                               ;;         "COMPLETED"
                                               ;;         (car all-curr-states))))
                            (newstatus            (if (or (> bad-not-started 0)
							  (and (equal? newstate "NOT_STARTED")
							       (> num-non-completes 0)))
						      "CHECK"
						      "STARTED"
                                                      (car all-curr-statuses))))
                       ;; (print "bad-not-supported: " bad-not-support " all-curr-states: " all-curr-states " all-curr-statuses: " all-curr-states)
                       ;;      " newstate: " newstate " newstatus: " newstatus)
                       ;; NB// Pass the db so it is part of the transaction
                       (db:test-set-state-status db run-id tl-test-id newstate newstatus #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)))))

;; BBnote: db:get-all-state-status-counts-for-test returns dbr:counts object aggregating state and status of items of a given test, *not including rollup state/status*
(define (db:get-all-state-status-counts-for-test dbstruct run-id test-name item-path)
  (db:with-db
   dbstruct #f #f
   (lambda (db)
     (sqlite3:map-row
      (lambda (state status count)
	(make-dbr:counts state: state status: status count: count))

;; Note: mode 'normal means that tests must be COMPLETED and ok (i.e. PASS, WARN, CHECK, SKIP or WAIVED)
;;       mode 'toplevel means that tests must be COMPLETED only
;;       mode 'itemmatch or 'itemwait means that tests items must be COMPLETED and (PASS|WARN|WAIVED|CHECK) [[ NB// NOT IMPLEMENTED YET ]]
;;       mode 'exclusive means this test/item cannot run if the same test/item is LAUNCHED,REMOTEHOSTSTART or RUNNING
;; 
;; (define (db:get-prereqs-not-met dbstruct run-id waitons ref-item-path mode)
(define (db:get-prereqs-not-met dbstruct run-id waitons ref-test-name ref-item-path mode itemmaps) ;; #!key (mode '(normal))(itemmap #f))
  ;; BBnote - rollup of an itemized test's overall state/status done in db:set-state-status-and-roll-up-items
  (append
   (if (member 'exclusive mode)
       (let ((running-tests (db:get-tests-for-run dbstruct
						  #f  ;; run-id of #f means for all runs. 
						  (if (string=? ref-item-path "")   ;; testpatt
						      ref-test-name
						      (conc ref-test-name "/" ref-item-path))

	    ;; and related sub items
	    ;; next should be using mt:get-tests-for-run?
	    (let ((tests             (db:get-tests-for-run-state-status dbstruct run-id waitontest-name))
		  (ever-seen         #f)
		  (parent-waiton-met #f)
		  (item-waiton-met   #f))
	      (for-each 
	       (lambda (test)
	       (lambda (test) ;; BB- this is the upstream test
		 ;; (if (equal? waitontest-name (db:test-get-testname test)) ;; by defintion this had better be true ...
		 (let* ((state             (db:test-get-state test))
			(status            (db:test-get-status test))
			(item-path         (db:test-get-item-path test))
			(item-path         (db:test-get-item-path test)) ;; BB- this is the upstream itempath
			(is-completed      (equal? state "COMPLETED"))
			(is-running        (equal? state "RUNNING"))
			(is-killed         (equal? state "KILLED"))
			(is-ok             (member status '("PASS" "WARN" "CHECK" "WAIVED" "SKIP")))
			;;                                       testname-b    path-a    path-b
			(same-itempath     (db:compare-itempaths ref-test-name item-path ref-item-path itemmaps))) ;; (equal? ref-item-path item-path)))
		   (set! ever-seen #t)

		     (set! parent-waiton-met #t))
		    ;; For itemwait mode IFF the previous matching item is good the set parent-waiton-met
		    ((and (not (null? (lset-intersection eq? mode '(itemmatch itemwait)))) ;; how is that different from (member mode '(itemmatch itemwait)) ?????
			  ;; (not (equal? item-path "")) ;; this applies to both top level (to allow launching of next batch) and items
			  same-itempath)
		     (if (and is-completed is-ok)
			 (set! item-waiton-met #t))
		     (if (and (equal? item-path "")
			      (or is-completed is-running));; this is the parent, set it to run if completed or running
		     (if (and (equal? item-path "") ;; if upstream rollup test is completed, parent-waiton-met is set
			      (or is-completed is-running));; this is the parent, set it to run if completed or running ;; BB1
			 (set! parent-waiton-met #t)))
		    ;; normal checking of parent items, any parent or parent item not ok blocks running
		    ((and is-completed
			  (or is-ok 
			      (member 'toplevel mode))              ;; toplevel does not block on FAIL
			  (and is-ok (member 'itemmatch mode))) ;; itemmatch blocks on not ok
		     (set! item-waiton-met #t)))))

(use regex typed-records matchable)

(declare (unit dcommon))

(declare (uses megatest-version))
(declare (uses gutils))
(declare (uses db))
(declare (uses synchash))
;; (declare (uses synchash))

(include "common_records.scm")
(include "db_records.scm")
(include "key_records.scm")
(include "run_records.scm")

;; yes, this is non-ideal 

;;  1. Make "data" hash-table hierarchial store of all displayed data
;;  2. Update synchash to understand "get-runs", "get-tests" etc.
;;  3. Add extraction of filters to synchash calls
;;
;;    NOTE: Used in newdashboard
;;
;; Mode is 'full or 'incremental for full refresh or incremental refresh
(define (dcommon:run-update keys data runname keypatts testpatt states statuses mode window-id)
  (let* (;; count and offset => #f so not used
	 ;; the synchash calls modify the "data" hash
	 (changed         #f)
	 (get-runs-sig    (conc (client:get-signature) " get-runs"))
	 (get-tests-sig   (conc (client:get-signature) " get-tests"))
	 (get-details-sig (conc (client:get-signature) " get-test-details"))

	 ;; test-ids to get and display are indexed on window-id in curr-test-ids hash
	 (test-ids        (hash-table-values (dboard:tabdat-curr-test-ids data)))
	 ;; run-id is #f in next line to send the query to server 0
 	 (run-changes     (synchash:client-get 'db:get-runs get-runs-sig (length keypatts) data #f runname #f #f keypatts))
	 (tests-detail-changes (if (not (null? test-ids))
				   (synchash:client-get 'db:get-test-info-by-ids get-details-sig 0  data #f test-ids)
				   '()))

	 ;; Now can calculate the run-ids
	 (run-hash    (hash-table-ref/default data get-runs-sig #f))
	 (run-ids     (if run-hash (filter number? (hash-table-keys run-hash)) '()))

	 (all-test-changes (let ((res (make-hash-table)))
			     (for-each (lambda (run-id)
					 (if (> run-id 0)
					     (hash-table-set! res run-id (synchash:client-get 'db:get-tests-for-run-mindata get-tests-sig 0 data run-id 1 testpatt states statuses #f))))
				       run-ids)
			     res))
	 (runs-hash    (hash-table-ref/default data get-runs-sig #f))
	 (header       (hash-table-ref/default runs-hash "header" #f))
	 (run-ids      (sort (filter number? (hash-table-keys runs-hash))
			     (lambda (a b)
			       (let* ((record-a (hash-table-ref runs-hash a))
				      (record-b (hash-table-ref runs-hash b))
				      (time-a   (db:get-value-by-header record-a header "event_time"))
				      (time-b   (db:get-value-by-header record-b header "event_time")))
				 (> time-a time-b)))
			     ))
	 (runid-to-col    (hash-table-ref *cachedata* "runid-to-col"))
	 (testname-to-row (hash-table-ref *cachedata* "testname-to-row")) 
	 (colnum       1)
	 (rownum       0)
	 (cellname (conc rownum ":" colnum))) ;; rownum = 0 is the header
;; (debug:print 0 *default-log-port* "test-ids " test-ids ", tests-detail-changes " tests-detail-changes)
    
	 ;; tests related stuff
	 ;; (all-testnames (delete-duplicates (map db:test-get-testname test-changes))))

    ;; Given a run-id and testname/item_path calculate a cell R:C

    ;; NOTE: Also build the test tree browser and look up table
    ;;
    ;; Each run is unique on its keys and runname or run-id, store in hash on colnum
    (for-each (lambda (run-id)
		(let* ((run-record (hash-table-ref/default runs-hash run-id #f))
		       (key-vals   (map (lambda (key)(db:get-value-by-header run-record header key))
					keys))
		       (run-name   (db:get-value-by-header run-record header "runname"))
		       (col-name   (conc (string-intersperse key-vals "\n") "\n" run-name))
		       (run-path   (append key-vals (list run-name))))
		  (hash-table-set! (dboard:tabdat-run-keys data) run-id run-path)
		  ;; modify cell - but only if changed
		  (set! changed (dcommon:modifiy-if-different (dboard:tabdat-runs-matrix data) cellname col-name changed))
		  (hash-table-set! runid-to-col run-id (list colnum run-record))
		  ;; Here we update the tests treebox and tree keys
		  (tree:add-node (dboard:tabdat-tests-tree data) "Runs" (append key-vals (list run-name))
				 userdata: (conc "run-id: " run-id))
		  (set! colnum (+ colnum 1))))
	      run-ids)

    ;; Scan all tests to be displayed and organise all the test names, respecting what is in the hash table
    ;; Do this analysis in the order of the run-ids, the most recent run wins
    (for-each (lambda (run-id)
		(let* ((run-path       (hash-table-ref (dboard:tabdat-run-keys data) run-id))
		       (test-changes   (hash-table-ref all-test-changes run-id))
		       (new-test-dat   (car test-changes))
		       (removed-tests  (cadr test-changes))
		       (tests          (sort (map cadr (filter (lambda (testrec)
								 (eq? run-id (db:mintest-get-run_id (cadr testrec))))
							       new-test-dat))
					     (lambda (a b)
					       (let ((time-a (db:mintest-get-event_time a))
						     (time-b (db:mintest-get-event_time b)))
						 (> time-a time-b)))))
		       ;; test-changes is a list of (( id record ) ... )
		       ;; Get list of test names sorted by time, remove tests
		       (test-names (delete-duplicates (map (lambda (t)
							     (let ((i (db:mintest-get-item_path t))
								   (n (db:mintest-get-testname  t)))
							       (if (string=? i "")
								   (conc "   " i)
								   n)))
							   tests)))
		       (colnum     (car (hash-table-ref runid-to-col run-id))))
		  ;; for each test name get the slot if it exists and fill in the cell
		  ;; or take the next slot and fill in the cell, deal with items in the
		  ;; run view panel? The run view panel can have a tree selector for
		  ;; browsing the tests/items

		  ;; SWITCH THIS TO USING CHANGED TESTS ONLY
		  (for-each (lambda (test)
			      (let* ((test-id   (db:mintest-get-id test))
				     (state     (db:mintest-get-state test))
				     (status    (db:mintest-get-status test))
				     (testname  (db:mintest-get-testname test))
				     (itempath  (db:mintest-get-item_path test))
				     (fullname  (conc testname "/" itempath))
				     (dispname  (if (string=? itempath "") testname (conc "   " itempath)))
				     (rownum    (hash-table-ref/default testname-to-row fullname #f))
				     (test-path (append run-path (if (equal? itempath "") 
								     (list testname)
								     (list testname itempath))))
				     (tb         (dboard:tabdat-tests-tree data)))
				(print "INFONOTE: run-path: " run-path)
				(tree:add-node (dboard:tabdat-tests-tree data) "Runs" 
					       test-path
					       userdata: (conc "test-id: " test-id))
				(let ((node-num (tree:find-node tb (cons "Runs" test-path)))
				      (color    (car (gutils:get-color-for-state-status state status))))
				  (debug:print 0 *default-log-port* "node-num: " node-num ", color: " color)

				  (set! changed (dcommon:modifiy-if-different 
						 tb
						 (conc "COLOR" node-num)
						 color changed))

				  ;; (iup:attribute-set! tb (conc "COLOR" node-num) color)
;; (define (dcommon:run-update keys data runname keypatts testpatt states statuses mode window-id)
;;   (let* (;; count and offset => #f so not used
;; 	 ;; the synchash calls modify the "data" hash
;; 	 (changed         #f)
;; 	 (get-runs-sig    (conc (client:get-signature) " get-runs"))
;; 	 (get-tests-sig   (conc (client:get-signature) " get-tests"))
;; 	 (get-details-sig (conc (client:get-signature) " get-test-details"))
;; 
;; 	 ;; test-ids to get and display are indexed on window-id in curr-test-ids hash
;; 	 (test-ids        (hash-table-values (dboard:tabdat-curr-test-ids data)))
;; 	 ;; run-id is #f in next line to send the query to server 0
;;  	 (run-changes     (synchash:client-get 'db:get-runs get-runs-sig (length keypatts) data #f runname #f #f keypatts))
;; 	 (tests-detail-changes (if (not (null? test-ids))
;; 				   (synchash:client-get 'db:get-test-info-by-ids get-details-sig 0  data #f test-ids)
;; 				   '()))
;; 
;; 	 ;; Now can calculate the run-ids
;; 	 (run-hash    (hash-table-ref/default data get-runs-sig #f))
;; 	 (run-ids     (if run-hash (filter number? (hash-table-keys run-hash)) '()))
;; 
;; 	 (all-test-changes (let ((res (make-hash-table)))
;; 			     (for-each (lambda (run-id)
;; 					 (if (> run-id 0)
;; 					     (hash-table-set! res run-id (synchash:client-get 'db:get-tests-for-run-mindata get-tests-sig 0 data run-id 1 testpatt states statuses #f))))
;; 				       run-ids)
;; 			     res))
;; 	 (runs-hash    (hash-table-ref/default data get-runs-sig #f))
;; 	 (header       (hash-table-ref/default runs-hash "header" #f))
;; 	 (run-ids      (sort (filter number? (hash-table-keys runs-hash))
;; 			     (lambda (a b)
;; 			       (let* ((record-a (hash-table-ref runs-hash a))
;; 				      (record-b (hash-table-ref runs-hash b))
;; 				      (time-a   (db:get-value-by-header record-a header "event_time"))
;; 				      (time-b   (db:get-value-by-header record-b header "event_time")))
;; 				 (> time-a time-b)))
;; 			     ))
;; 	 (runid-to-col    (hash-table-ref *cachedata* "runid-to-col"))
;; 	 (testname-to-row (hash-table-ref *cachedata* "testname-to-row")) 
;; 	 (colnum       1)
;; 	 (rownum       0)
;; 	 (cellname (conc rownum ":" colnum))) ;; rownum = 0 is the header
;; ;; (debug:print 0 *default-log-port* "test-ids " test-ids ", tests-detail-changes " tests-detail-changes)
;;     
;; 	 ;; tests related stuff
;; 	 ;; (all-testnames (delete-duplicates (map db:test-get-testname test-changes))))
;; 
;;     ;; Given a run-id and testname/item_path calculate a cell R:C
;; 
;;     ;; NOTE: Also build the test tree browser and look up table
;;     ;;
;;     ;; Each run is unique on its keys and runname or run-id, store in hash on colnum
;;     (for-each (lambda (run-id)
;; 		(let* ((run-record (hash-table-ref/default runs-hash run-id #f))
;; 		       (key-vals   (map (lambda (key)(db:get-value-by-header run-record header key))
;; 					keys))
;; 		       (run-name   (db:get-value-by-header run-record header "runname"))
;; 		       (col-name   (conc (string-intersperse key-vals "\n") "\n" run-name))
;; 		       (run-path   (append key-vals (list run-name))))
;; 		  (hash-table-set! (dboard:tabdat-run-keys data) run-id run-path)
;; 		  ;; modify cell - but only if changed
;; 		  (set! changed (dcommon:modifiy-if-different (dboard:tabdat-runs-matrix data) cellname col-name changed))
;; 		  (hash-table-set! runid-to-col run-id (list colnum run-record))
;; 		  ;; Here we update the tests treebox and tree keys
;; 		  (tree:add-node (dboard:tabdat-tests-tree data) "Runs" (append key-vals (list run-name))
;; 				 userdata: (conc "run-id: " run-id))
;; 		  (set! colnum (+ colnum 1))))
;; 	      run-ids)
;; 
;;     ;; Scan all tests to be displayed and organise all the test names, respecting what is in the hash table
;;     ;; Do this analysis in the order of the run-ids, the most recent run wins
;;     (for-each (lambda (run-id)
;; 		(let* ((run-path       (hash-table-ref (dboard:tabdat-run-keys data) run-id))
;; 		       (test-changes   (hash-table-ref all-test-changes run-id))
;; 		       (new-test-dat   (car test-changes))
;; 		       (removed-tests  (cadr test-changes))
;; 		       (tests          (sort (map cadr (filter (lambda (testrec)
;; 								 (eq? run-id (db:mintest-get-run_id (cadr testrec))))
;; 							       new-test-dat))
;; 					     (lambda (a b)
;; 					       (let ((time-a (db:mintest-get-event_time a))
;; 						     (time-b (db:mintest-get-event_time b)))
;; 						 (> time-a time-b)))))
;; 		       ;; test-changes is a list of (( id record ) ... )
;; 		       ;; Get list of test names sorted by time, remove tests
;; 		       (test-names (delete-duplicates (map (lambda (t)
;; 							     (let ((i (db:mintest-get-item_path t))
;; 								   (n (db:mintest-get-testname  t)))
;; 							       (if (string=? i "")
;; 								   (conc "   " i)
;; 								   n)))
;; 							   tests)))
;; 		       (colnum     (car (hash-table-ref runid-to-col run-id))))
;; 		  ;; for each test name get the slot if it exists and fill in the cell
;; 		  ;; or take the next slot and fill in the cell, deal with items in the
;; 		  ;; run view panel? The run view panel can have a tree selector for
;; 		  ;; browsing the tests/items
;; 
;; 		  ;; SWITCH THIS TO USING CHANGED TESTS ONLY
;; 		  (for-each (lambda (test)
;; 			      (let* ((test-id   (db:mintest-get-id test))
;; 				     (state     (db:mintest-get-state test))
;; 				     (status    (db:mintest-get-status test))
;; 				     (testname  (db:mintest-get-testname test))
;; 				     (itempath  (db:mintest-get-item_path test))
;; 				     (fullname  (conc testname "/" itempath))
;; 				     (dispname  (if (string=? itempath "") testname (conc "   " itempath)))
;; 				     (rownum    (hash-table-ref/default testname-to-row fullname #f))
;; 				     (test-path (append run-path (if (equal? itempath "") 
;; 								     (list testname)
;; 								     (list testname itempath))))
;; 				     (tb         (dboard:tabdat-tests-tree data)))
;; 				(print "INFONOTE: run-path: " run-path)
;; 				(tree:add-node (dboard:tabdat-tests-tree data) "Runs" 
;; 					       test-path
;; 					       userdata: (conc "test-id: " test-id))
;; 				(let ((node-num (tree:find-node tb (cons "Runs" test-path)))
;; 				      (color    (car (gutils:get-color-for-state-status state status))))
;; 				  (debug:print 0 *default-log-port* "node-num: " node-num ", color: " color)
;; 
;; 				  (set! changed (dcommon:modifiy-if-different 
;; 						 tb
;; 						 (conc "COLOR" node-num)
;; 						 color changed))
;; 
;; 				  ;; (iup:attribute-set! tb (conc "COLOR" node-num) color)
				  )
				(hash-table-set! (dboard:tabdat-path-test-ids data) test-path test-id)
				(if (not rownum)
				    (let ((rownums (hash-table-values testname-to-row)))
				      (set! rownum (if (null? rownums)
;; 				  )
;; 				(hash-table-set! (dboard:tabdat-path-test-ids data) test-path test-id)
;; 				(if (not rownum)
;; 				    (let ((rownums (hash-table-values testname-to-row)))
;; 				      (set! rownum (if (null? rownums)
						       1
						       (+ 1 (common:max rownums))))
				      (hash-table-set! testname-to-row fullname rownum)
				      ;; create the label
				      (set! changed (dcommon:modifiy-if-different 
						     (dboard:tabdat-runs-matrix data)
						     (conc rownum ":" 0)
						     dispname
						     changed))
				      ;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
				      ;;   		  (conc rownum ":" 0) dispname)
				      ))
				;; set the cell text and color
				;; (debug:print 2 *default-log-port* "rownum:colnum=" rownum ":" colnum ", state=" status)
				(set! changed (dcommon:modifiy-if-different 
						     (dboard:tabdat-runs-matrix data)
						     (conc rownum ":" colnum)
						     (if (member state '("ARCHIVED" "COMPLETED"))
							 status
							 state)
						     changed))
				;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
				;; 		    (conc rownum ":" colnum)
				;; 		    (if (member state '("ARCHIVED" "COMPLETED"))
				;; 			status
				;; 			state))
				(set! changed (dcommon:modifiy-if-different 
					       (dboard:tabdat-runs-matrix data)
					       (conc "BGCOLOR" rownum ":" colnum)
					       (car (gutils:get-color-for-state-status state status))
					       changed))
				;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
				;; 		    (conc "BGCOLOR" rownum ":" colnum)
				;; 		    (car (gutils:get-color-for-state-status state status)))
				))
			    tests)))
	      run-ids)

    (let ((updater (hash-table-ref/default  (dboard:commondat-updaters commondat) window-id #f)))
      (if updater (updater (hash-table-ref/default data get-details-sig #f))))

    (if changed (iup:attribute-set! (dboard:tabdat-runs-matrix data) "REDRAW" "ALL"))
    ;; (debug:print 2 *default-log-port* "run-changes: " run-changes)
    ;; (debug:print 2 *default-log-port* "test-changes: " test-changes)
    (list run-changes all-test-changes)))
;; 						       1
;; 						       (+ 1 (common:max rownums))))
;; 				      (hash-table-set! testname-to-row fullname rownum)
;; 				      ;; create the label
;; 				      (set! changed (dcommon:modifiy-if-different 
;; 						     (dboard:tabdat-runs-matrix data)
;; 						     (conc rownum ":" 0)
;; 						     dispname
;; 						     changed))
;; 				      ;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
;; 				      ;;   		  (conc rownum ":" 0) dispname)
;; 				      ))
;; 				;; set the cell text and color
;; 				;; (debug:print 2 *default-log-port* "rownum:colnum=" rownum ":" colnum ", state=" status)
;; 				(set! changed (dcommon:modifiy-if-different 
;; 						     (dboard:tabdat-runs-matrix data)
;; 						     (conc rownum ":" colnum)
;; 						     (if (member state '("ARCHIVED" "COMPLETED"))
;; 							 status
;; 							 state)
;; 						     changed))
;; 				;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
;; 				;; 		    (conc rownum ":" colnum)
;; 				;; 		    (if (member state '("ARCHIVED" "COMPLETED"))
;; 				;; 			status
;; 				;; 			state))
;; 				(set! changed (dcommon:modifiy-if-different 
;; 					       (dboard:tabdat-runs-matrix data)
;; 					       (conc "BGCOLOR" rownum ":" colnum)
;; 					       (car (gutils:get-color-for-state-status state status))
;; 					       changed))
;; 				;; (iup:attribute-set! (dboard:tabdat-runs-matrix data)
;; 				;; 		    (conc "BGCOLOR" rownum ":" colnum)
;; 				;; 		    (car (gutils:get-color-for-state-status state status)))
;; 				))
;; 			    tests)))
;; 	      run-ids)
;; 
;;     (let ((updater (hash-table-ref/default  (dboard:commondat-updaters commondat) window-id #f)))
;;       (if updater (updater (hash-table-ref/default data get-details-sig #f))))
;; 
;;     (if changed (iup:attribute-set! (dboard:tabdat-runs-matrix data) "REDRAW" "ALL"))
;;     ;; (debug:print 2 *default-log-port* "run-changes: " run-changes)
;;     ;; (debug:print 2 *default-log-port* "test-changes: " test-changes)
;;     (list run-changes all-test-changes)))

(define (dcommon:runsdat-get-col-num dat target runname force-set)
  (let* ((runs-index (dboard:runsdat-runs-index dat))
	 (col-name   (conc target "/" runname))
	 (res        (hash-table-ref/default runs-index col-name #f)))
    (if res
	res

integrating and or running a complex suite of tests for release
qualification.</p></div>
</div>
</div>
<div class="sect1">
<h2 id="_megatest_design_philosophy">Megatest Design Philosophy</h2>
<div class="sectionbody">
<div class="paragraph"><p>Megatest is intended to provide the minimum needed resources to make
writing a suite of tests and tasks for implementing continuous build
for software, design engineering or process control (via owlfs for
example) without being specialized for any specific problem
space. Megatest in of itself does not know what constitutes a PASS or
FAIL of a test or task. In most cases megatest is best used in
conjunction with logpro or a similar tool to parse, analyze and decide
<div class="paragraph"><p>Megatest is a distributed system intended to provide the minimum needed
resources to make writing a suite of tests and tasks for implementing
continuous build for software, design engineering or process control (via
owlfs for example) without being specialized for any specific problem
space. Megatest in of itself does not know what constitutes a PASS or FAIL
of a test or task. In most cases megatest is best used in conjunction with
logpro or a similar tool to parse, analyze and decide on the test outcome.</p></div>
on the test outcome.</p></div>
<div class="ulist"><ul>
<li>
<p>
Self-checking -Repeatable strive for directed or self-checking test
   as opposed to delta based tests
</p>
</li>

<h3 id="_architecture_refactor">Architecture Refactor</h3>
<div class="sect3">
<h4 id="_goals">Goals</h4>
<div class="olist arabic"><ol class="arabic">
<li>
<p>
Reduce load on the file system. Sqlite3 files on network filesystem can be
  a burden.
  a burden. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Reduce number of servers and frequency of start/stop. This is mostly an
  issue of clutter but also a reduction in "moving parts".
  issue of clutter but also a reduction in "moving parts". <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Coalesce activities to a single home host where possible. Give the user
  feedback that they have started the dashboard on a host other than the
  home host.
  home host. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Reduce number of processes involved in managing running tests.
</p>
</li>
</ol></div>
</div>
<div class="sect3">
<h4 id="_changes_needed">Changes Needed</h4>
<div class="olist arabic"><ol class="arabic">
<li>
<p>
ACID compliant db will be on /tmp and synced to megatest.db with a five
  second max delay.
  second max delay. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Read/writes to db for processes on homehost will go direct to /tmp
  megatest.db file.
  megatest.db file. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Read/wites fron non-homehost processes will go through one server. Bulk
  reads (e.g. for dashboard or list-runs) will be cached on the current host
  in /tmp and synced from the home megatest.db in the testsuite area.
  in /tmp and synced from the home megatest.db in the testsuite area. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Db syncs rely on the target db file timestame minus some margin.
Db syncs rely on the target db file timestame minus some margin. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Since bulk reads do not use the server we can switch to simple RPC for the
  network transport.
  network transport. <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Test running manager process extended to manage multiple running tests.
</p>
</li>
</ol></div>
</div>
</div>
<div class="sect2">
<h3 id="_current_items">Current Items</h3>
<div class="sect3">
<h4 id="_ww05_migrate_to_inmem_db">ww05 - migrate to inmem-db</h4>
<div class="olist arabic"><ol class="arabic">
<li>
<p>
Switch to inmem db with fast sync to on disk db&#8217;s [DONE]
Switch to inmem db with fast sync to on disk db&#8217;s <span class="green">[DONE]</span>
</p>
</li>
<li>
<p>
Server polls tasks table for next action
</p>
<div class="olist loweralpha"><ol class="loweralpha">
<li>
<p>
Task table used for tracking runner process [DONE]
Task table used for tracking runner process <span class="red">[Replaced by mtutil]</span>
</p>
</li>
<li>
<p>
Task table used for jobs to run
Task table used for jobs to run <span class="red">[Replaced by mtutil]</span>
</p>
</li>
<li>
<p>
Task table used for queueing runner actions (remove runs, cleanRunExecute, etc)
Task table used for queueing runner actions (remove runs,
   cleanRunExecute, etc)  <span class="red">[Replaced by mtutil</span>]
</p>
</li>
</ol></div>
</li>
</ol></div>
<div class="paragraph"><p>shifting, note that the preceding blank line is needed.</p></div>
</div>

<div class="listingblock">
<div class="content monospaced">
<pre>[items]
A a b c
B d e f</pre>
</div></div>
<div class="paragraph"><p>Then the config file would effectively appear to contain an items section
exactly like the output from the script. This is extremely useful when
dynamically creating items, itemstables and other config structures. You can
see the expansion of the call by looking in the cached files (look in your
linktree for megatest.config and runconfigs.config cache files and in your
test run areas for the expanded and cached testconfig).</p></div>
exactly like the output from the script. This is useful when dynamically
creating items, itemstables and other config structures. You can see the
expansion of the call by looking in the cached files (look in your linktree
for megatest.config and runconfigs.config cache files and in your test run
areas for the expanded and cached testconfig).</p></div>
<div class="paragraph"><p>Wildcards and regexes in Targets</p></div>
<div class="listingblock">
<div class="content monospaced">
<pre>[a/2/b]
VAR1 VAL1

[a/%/b]
VAR1 VAL2</pre>
</div></div>
<div class="paragraph"><p>Will result in:</p></div>
<div class="listingblock">
<div class="content monospaced">
<pre>[a/2/b]
VAR1 VAL2</pre>
</div></div>
<div class="paragraph"><p>Can use either wildcard of "%" or a regular expression:</p></div>
<div class="listingblock">
<div class="content monospaced">
<pre>[/abc.*def/]</pre>
</div></div>
<div class="sect3">
<h4 id="_disk_space_checks">Disk Space Checks</h4>
<div class="paragraph"><p>Some parameters you can put in the [setup] section of megatest.config:</p></div>
<div class="listingblock">
<div class="content monospaced">
<pre># minimum space required in a run disk
minspace 10000000

<table><tr>
<td class="icon">
<img src="/nfs/pdx/disks/ice.disk.002/icfadm/pkgs/asciidoc/8.6.7/images/icons/note.png" alt="Note">
</td>
<td class="content">There is a trailing space after the --</td>
</tr></table>
</div>
<div class="paragraph"><p>There are a number of environment variables available to the trigger script
but since triggers can be called in various contexts not all variables are
available at all times. The trigger script should check for the variable and
fail gracefully if it doesn&#8217;t exist.</p></div>
<table class="tableblock frame-topbot grid-all"
style="
width:90%;
">
<caption class="title">Table 4. Environment variables visible to the trigger script</caption>
<col style="width:33%;">
<col style="width:66%;">
<thead>
<tr>
<th class="tableblock halign-center valign-top" >Variable             </th>
<th class="tableblock halign-left valign-top" > Purpose</th>
</tr>
</thead>
<tbody>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_TEST_RUN_DIR</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The directory where Megatest ran this test</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_CMDINFO</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">Encoded command data for the test</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_DEBUG_MODE</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">Used to pass the debug mode to nested calls to Megatest</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_RUN_AREA_HOME</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">Megatest home area</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_TESTSUITENAME</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The name of this testsuite or area</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_TEST_NAME</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The name of this test</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_ITEM_INFO</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The variable and values for the test item</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_MEGATEST</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">Which Megatest binary is being used by this area</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_TARGET</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The target variable values, separated by <em>/</em></p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_LINKTREE</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The base of the link tree where all run tests can be found</p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_ITEMPATH</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The values of the item path variables, separated by <em>/</em></p></td>
</tr>
<tr>
<td class="tableblock halign-center valign-top" ><p class="tableblock">MT_RUNNAME</p></td>
<td class="tableblock halign-left valign-top" ><p class="tableblock monospaced">The name of the run</p></td>
</tr>
</tbody>
</table>
</div>
<div class="sect2">
<h3 id="_override_the_toplevel_html_file">Override the Toplevel HTML File</h3>
<div class="paragraph"><p>Megatest generates a simple html file summary for top level tests of
iterated tests. The generation can be overridden. NOTE: the output of
the script is captured from stdout to create the html.</p></div>
<div class="listingblock">

<h2 id="_programming_api">Programming API</h2>
<div class="sectionbody">
<div class="paragraph"><p>These routines can be called from the megatest repl.</p></div>
<table class="tableblock frame-topbot grid-all"
style="
width:70%;
">
<caption class="title">Table 4. API Keys Related Calls</caption>
<caption class="title">Table 5. API Keys Related Calls</caption>
<col style="width:14%;">
<col style="width:28%;">
<col style="width:28%;">
<col style="width:28%;">
<thead>
<tr>
<th class="tableblock halign-center valign-top" >API Call                        </th>

</div>
</div>
</div>
<div id="footnotes"><hr></div>
<div id="footer">
<div id="footer-text">
Version 1.0<br>
Last updated 2016-12-12 13:03:08 PST
Last updated 2017-05-10 08:25:49 PDT
</div>
</div>
</body>
</html>

tool, flexible enough to meet the needs of any team doing continuous
integrating and or running a complex suite of tests for release
qualification.

Megatest Design Philosophy
--------------------------

Megatest is intended to provide the minimum needed resources to make
writing a suite of tests and tasks for implementing continuous build
for software, design engineering or process control (via owlfs for
example) without being specialized for any specific problem
space. Megatest in of itself does not know what constitutes a PASS or
FAIL of a test or task. In most cases megatest is best used in
conjunction with logpro or a similar tool to parse, analyze and decide
Megatest is a distributed system intended to provide the minimum needed
resources to make writing a suite of tests and tasks for implementing
continuous build for software, design engineering or process control (via
owlfs for example) without being specialized for any specific problem
space. Megatest in of itself does not know what constitutes a PASS or FAIL
of a test or task. In most cases megatest is best used in conjunction with
logpro or a similar tool to parse, analyze and decide on the test outcome.
on the test outcome. 

 * Self-checking -Repeatable strive for directed or self-checking test
   as opposed to delta based tests

 * Traceable - environment variables, host OS and other possibly influential
   variables are captured and kept recorded.

-------------------------
[a/2/b]
VAR1 VAL2
-------------------------

Can use either wildcard of "%" or a regular expression:

-------------------------
[/abc.*def/]
-------------------------

Disk Space Checks
^^^^^^^^^^^^^^^^^

Some parameters you can put in the [setup] section of megatest.config:

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

-----------------
[triggers]
COMPLETED/ xterm -e bash -s -- 
-----------------

NOTE: There is a trailing space after the --

There are a number of environment variables available to the trigger script
but since triggers can be called in various contexts not all variables are
available at all times. The trigger script should check for the variable and
fail gracefully if it doesn't exist.

.Environment variables visible to the trigger script
[width="90%",cols="^,2m",frame="topbot",options="header"]
|======================
|Variable             | Purpose            
| MT_TEST_RUN_DIR     | The directory where Megatest ran this test                   
| MT_CMDINFO          | Encoded command data for the test                   
| MT_DEBUG_MODE       | Used to pass the debug mode to nested calls to Megatest                   
| MT_RUN_AREA_HOME    | Megatest home area 
| MT_TESTSUITENAME    | The name of this testsuite or area                   
| MT_TEST_NAME        | The name of this test
| MT_ITEM_INFO        | The variable and values for the test item
| MT_MEGATEST         | Which Megatest binary is being used by this area
| MT_TARGET           | The target variable values, separated by '/'
| MT_LINKTREE         | The base of the link tree where all run tests can be found
| MT_ITEMPATH         | The values of the item path variables, separated by '/'
| MT_RUNNAME          | The name of the run
|======================


Override the Toplevel HTML File
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Megatest generates a simple html file summary for top level tests of
iterated tests. The generation can be overridden. NOTE: the output of

Road Map
--------

Note 1: This road-map is still evolving and subject to change without notice.

Architecture Refactor
~~~~~~~~~~~~~~~~~~~~~

Goals
^^^^^

. Reduce load on the file system. Sqlite3 files on network filesystem can be
  a burden.
  a burden. [green]#[DONE]#
. Reduce number of servers and frequency of start/stop. This is mostly an
  issue of clutter but also a reduction in "moving parts".
  issue of clutter but also a reduction in "moving parts". [green]#[DONE]#
. Coalesce activities to a single home host where possible. Give the user
  feedback that they have started the dashboard on a host other than the
  home host.
  home host. [green]#[DONE]#
. Reduce number of processes involved in managing running tests.

Changes Needed
^^^^^^^^^^^^^^

. ACID compliant db will be on /tmp and synced to megatest.db with a five
  second max delay.
  second max delay. [green]#[DONE]#
. Read/writes to db for processes on homehost will go direct to /tmp
  megatest.db file.
  megatest.db file. [green]#[DONE]#
. Read/wites fron non-homehost processes will go through one server. Bulk
  reads (e.g. for dashboard or list-runs) will be cached on the current host
  in /tmp and synced from the home megatest.db in the testsuite area.
. Db syncs rely on the target db file timestame minus some margin.
  in /tmp and synced from the home megatest.db in the testsuite area. [green]#[DONE]#
. Db syncs rely on the target db file timestame minus some margin. [green]#[DONE]#
. Since bulk reads do not use the server we can switch to simple RPC for the
  network transport.
  network transport. [green]#[DONE]#
. Test running manager process extended to manage multiple running tests.

Current Items
~~~~~~~~~~~~~

ww05 - migrate to inmem-db
^^^^^^^^^^^^^^^^^^^^^^^^^^

. Switch to inmem db with fast sync to on disk db's [DONE]
. Switch to inmem db with fast sync to on disk db's [green]#[DONE]#
. Server polls tasks table for next action
.. Task table used for tracking runner process [DONE]
.. Task table used for jobs to run
.. Task table used for queueing runner actions (remove runs, cleanRunExecute, etc)
.. Task table used for tracking runner process [red]#[Replaced by mtutil]#
.. Task table used for jobs to run [red]#[Replaced by mtutil]#
.. Task table used for queueing runner actions (remove runs,
   cleanRunExecute, etc)  [red]#[Replaced by mtutil#]


// ww32
// ~~~~
// 
// . Rerun step and or subsequent steps from gui
// . Refresh test area files from gui

(use (prefix sqlite3 sqlite3:) posix typed-records) 

(define (open-tail-db )
  (let* ((basedir   (create-directory (conc "/tmp/" (current-user-name))))
	 (dbpath    (conc basedir "/megatest_logs.db"))
	 (dbexists  (file-exists? dbpath))
	 (db        (sqlite3:open-database dbpath))
	 (handler   (sqlite3:make-busy-timeout 136000)))
    (sqlite3:set-busy-handler! db handler)
    (sqlite3:execute db "PRAGMA synchronous = 0;")
    (if (not dbexists)
	(begin
	  (sqlite3:execute db "CREATE TABLE IF NOT EXISTS log_files (id INTEGER PRIMARY KEY,filename TEXT,event_time TIMESTAMP DEFAULT (strftime('%s','now')));")
	  (sqlite3:execute db "CREATE TABLE IF NOT EXISTS log_data  (id INTEGER PRIMARY KEY,fid INTEGER,line TEXT,event_time TIMESTAMP DEFAULT (strftime('%s','now')));")
	  ))
    db))

(define (tail-write db fid lines)
  (sqlite3:with-transaction
   db
   (lambda ()
     (for-each
      (lambda (line)
	(sqlite3:execute db "INSERT INTO log_data (fid,line) VALUES (?,?);" fid line))
      lines))))

(define (tail-get-fid db fname)
  (let ((fid   (handle-exceptions
		   exn
		   #f
		 (sqlite3:first-result db "SELECT id FROM log_files WHERE filename=?;" fname))))
    (if fid
	fid
	(begin
	  (sqlite3:execute db "INSERT INTO log_files (filename) VALUES (?);" fname)
	  (tail-get-fid db fname)))))

(define (file-tail fname #!key (db-in #f))
  (let* ((inp (open-input-file fname))
	 (db  (or db-in (open-tail-db)))
	 (fid (tail-get-fid db fname)))
    (let loop ((inl    (read-line inp))
	       (lines '())
	       (lastwr (current-seconds)))
      (if (eof-object? inl)
	  (let ((timed-out (> (- (current-seconds) lastwr) 60)))
	    (if timed-out (tail-write db fid (reverse lines)))
	    (sleep 1)
	    (if timed-out
		(loop (read-line inp) '() (current-seconds))
		(loop (read-line inp) lines lastwr)))
	  (let* ((savelines (> (length lines) 19)))
	    ;; (print inl)
	    (if savelines (tail-write db fid (reverse lines)))
	    (loop (read-line inp)
		  (if savelines
		      '()
		      (cons inl lines))
		  (if savelines
		      (current-seconds)
		      lastwr)))))))

;; offset -20 means get last 20 lines
;;
(define (tail-get-lines db fid offset count)
  (if (> offset 0)
      (map-row (lambda (id line)
		 (vector id line))
	       db
	       "SELECT id,line FROM log_data WHERE fid=? OFFSET ? LIMIT ?;" fid offset count)
      (reverse ;; get N from the end
       (map-row (lambda (id line)
		  (vector id line))
		db
		"SELECT id,line FROM log_data WHERE fid=? ORDER BY id DESC LIMIT ?;" fid (abs offset)))))

  '((PASS . "70 249 73")
    (FAIL . "253 33 49")
    (SKIP . "230 230 0")))

(define (gutils:get-color-spec effective-state)
  (or (alist-ref effective-state gutils:colors)
      (alist-ref 'FAIL gutils:colors)))
 

;; BBnote - state status dashboard button color / text defined here
(define (gutils:get-color-for-state-status state status);; #!key (get-label #f))
  ;; ((if get-label cadr car)
  (case (string->symbol state)
    ((COMPLETED) ;; ARCHIVED)
     (case (string->symbol status)
       ((PASS)        (list "70  249 73" status))
       ((WARN WAIVED) (list "255 172 13" status))

    ;;	  (if (or (equal? status "WARN")
    ;;		  (equal? status "WAIVED"))
    ;;	      (list "255 172 13" status)
    ;;	      (list "223 33 49"  status)))) ;; greenish orangeish redish
    ((LAUNCHED)         (list "101 123 142"  state))
    ((CHECK)            (list "255 100 50"   state))
    ((REMOTEHOSTSTART)  (list "50 130 195"   state))
    ((RUNNING)          (list "9 131 232"    state))
    ((RUNNING STARTED)          (list "9 131 232"    state))
    ((KILLREQ)          (list "39 82 206"    state))
    ((KILLED)           (list "234 101 17"   state))
    ((NOT_STARTED)      (case (string->symbol status)
			  ((CHECK)(list (gutils:get-color-spec 'SKIP) state))
			  ((CHECK STARTED)(list (gutils:get-color-spec 'SKIP) state))
			  (else   (list "240 240 240"                 state))))
    ;; for xor mode below
    ;;
    ((CLEAN)
     (case (string->symbol status)
       ((CLEAN-FAIL CLEAN-CHECK CLEAN-ABORT)  (list "200 130 13" status)) ;; orange requested for these
       (else  (list "60  235 63" status))))
    ((DIRTY-BETTER)     (list "160  255 153" status))
    ((DIRTY-WORSE)      (list "165 42  42" status))
    ((BOTH-BAD)         (list "180 33 49" status))

    (else               (list "192 192 192"  state))))

	exn
	(begin
	  (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 0 *default-log-port* "exn=" (condition->list 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")
		(thread-sleep! 0.1)
		
		;; get_next_port goes here
		(http-transport:try-start-server ipaddrstr
						 (portlogger:open-run-close portlogger:find-port)))

;;    - could be netbatch
;;      (launch-test db (cadr status) test-conf))
(define (launch-test test-id run-id run-info keyvals runname test-conf test-name test-path itemdat params)
  (mutex-lock! *launch-setup-mutex*) ;; setting variables and processing the testconfig is NOT thread-safe, reuse the launch-setup mutex
  (let* ((item-path       (item-list->path itemdat))
	 (contour         #f)) ;; NOT READY FOR THIS (args:get-arg "-contour")))
    (let loop ((delta        (- (current-seconds) *last-launch*))
	       (launch-delay (string->number (or (configf:lookup *configdat* "setup" "launch-delay") "0"))))
	       (launch-delay (configf:lookup-number *configdat* "setup" "launch-delay" default: 1)))
      (if (> launch-delay delta)
	  (begin
	    (debug:print-info 0 *default-log-port* "Delaying launch of " test-name " for " (- launch-delay delta) " seconds")
	    (thread-sleep! (- launch-delay delta))
	    (loop (- (current-seconds) *last-launch*) launch-delay))))
    (change-directory *toppath*)
    (alist->env-vars ;; consolidate this code with the code in megatest.scm for "-execute", *maybe* - the longer they are set the longer each launch takes (must be non-overlapping with the vars)

# [fields]
# a text
# b text
# c text

[defaults]
# control over usercode location not implemented, for now must be .mtutil.scm
usercode    .mtutil.scm
areafilter  area-to-run
targtrans   generic-target-translator
runtrans    generic-runname-translator

[setup]
pktsdirs /tmp/mt_pkts /some/other/source

		       
		     (for-each 
		      (lambda (test)
		      	(common:debug-handle-exceptions #f
			 exn
			 (begin
			   (debug:print-error 0 *default-log-port* "Bad data in test record? " test)
			   (print "exn=" (condition->list exn))
			   (debug:print-error 5 *default-log-port* "exn=" (condition->list exn))
			   (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
			   (print-call-chain (current-error-port)))
			 (let* ((test-id      (if (member "id"           tests-spec)(get-value-by-fieldname test test-field-index "id"          ) #f)) ;; (db:test-get-id         test))
				(testname     (if (member "testname"     tests-spec)(get-value-by-fieldname test test-field-index "testname"    ) #f)) ;; (db:test-get-testname   test))
				(itempath     (if (member "item_path"    tests-spec)(get-value-by-fieldname test test-field-index "item_path"   ) #f)) ;; (db:test-get-item-path  test))
				(comment      (if (member "comment"      tests-spec)(get-value-by-fieldname test test-field-index "comment"     ) #f)) ;; (db:test-get-comment    test))
				(tstate       (if (member "state"        tests-spec)(get-value-by-fieldname test test-field-index "state"       ) #f)) ;; (db:test-get-state      test))

(if (or (args:any? "-h" "help" "-help" "--help")
	(member *action* '("-h" "-help" "--help" "help")))
    (begin
      (print help)
      (exit 1)))

;;======================================================================
;; pkts
;;======================================================================

(define (with-queue-db mtconf proc)
  (let* ((pktsdirs (configf:lookup mtconf "setup"  "pktsdirs"))
	 (pktsdir  (if pktsdirs (car (string-split pktsdirs " ")) #f))
	 (toppath  (configf:lookup mtconf "dyndat" "toppath"))
	 (pdbpath  (or (configf:lookup mtconf "setup"  "pdbpath") pktsdir)))
    (if (not (and  pktsdir toppath pdbpath))
	(begin
	  (print "ERROR: settings are missing in your megatest.config for area management.")
	  (print "  you need to have pktsdir in the [setup] section."))
	(let* ((pdb  (open-queue-db pdbpath "pkts.db"
				    schema: '("CREATE TABLE groups (id INTEGER PRIMARY KEY,groupname TEXT, CONSTRAINT group_constraint UNIQUE (groupname));")))
	       (res  (proc pktsdirs pktsdir pdb)))
	  (dbi:close pdb)
	  res
	  ))))

(define (load-pkts-to-db mtconf)
  (with-queue-db
   mtconf
   (lambda (pktsdirs pktsdir pdb)
     (for-each
      (lambda (pktsdir) ;; look at all
	(if (and (file-exists? pktsdir)
		 (directory? pktsdir)
		 (file-read-access? pktsdir))
	    (let ((pkts (glob (conc pktsdir "/*.pkt"))))
	      (for-each
	       (lambda (pkt)
		 (let* ((uuid    (cadr (string-match ".*/([0-9a-f]+).pkt" pkt)))
			(exists  (lookup-by-uuid pdb uuid #f)))
		   (if (not exists)
		       (let* ((pktdat (string-intersperse
				       (with-input-from-file pkt read-lines)
				       "\n"))
			      (apkt   (pkt->alist pktdat))
			      (ptype  (alist-ref 'T apkt))
			      (parent (alist-ref 'P apkt)))
			 (add-to-queue pdb pktdat uuid (or ptype 'cmd) parent 0)
			 (debug:print 4 *default-log-port* "Added " uuid " of type " ptype " to queue"))
		       (debug:print 4 *default-log-port* "pkt: " uuid " exists, skipping...")
		       )))
	       pkts))))
      (string-split pktsdirs)))))

(define (get-pkt-alists pkts)
  (map (lambda (x)
	 (alist-ref 'apkt x)) ;; 'pkta pulls out the alist from the read pkt
       pkts))

;; given list of pkts (alist mode) return list of D cards as Unix epoch, sorted descending
;; also delete duplicates by target i.e. (car pkt)
(define (get-pkt-times pkts)
  (delete-duplicates
   (sort 
    (map (lambda (x)
	   `(,(alist-ref 't x) . ,(string->number (alist-ref 'D x))))
	 pkts)
    (lambda (a b)(> (cdr a)(cdr b))))      ;; sort descending
   (lambda (a b)(equal? (car a)(car b))))) ;; remove duplicates by target

;;======================================================================
;; Runs
;;======================================================================

;; make a runname
;;
(define (make-runname pre post)

	 (mtconfdat (find-and-read-config        ;; NB// sets MT_RUN_AREA_HOME as side effect
		     mtconfig
		     ;; environ-patt: "env-override"
		     given-toppath: start-dir
		     ;; pathenvvar: "MT_RUN_AREA_HOME"
		     ))
	 (mtconf    (if mtconfdat (car mtconfdat) #f)))
    ;; we set some dynamic data in a section called "dyndata"
    ;; we set some dynamic data in a section called "scratchdata"
    (if mtconf
	(begin
	  (configf:section-var-set! mtconf "dyndat" "toppath" start-dir)))
    ;; (print "TOPPATH: " (configf:lookup mtconf "dyndat" "toppath"))
	  (configf:section-var-set! mtconf "scratchdat" "toppath" start-dir)))
    ;; (print "TOPPATH: " (configf:lookup mtconf "scratchdat" "toppath"))
    mtconfdat))


;; NEED TIMESTAMP ON PKTS for efficient loading of packets into db.


;; make a run request pkt from basic data, this seriously needs to be refactored

;; (use trace)(trace create-run-pkt)

;; collect all needed data and create run pkts for contours with changed inputs
;;
(define (generate-run-pkts mtconf toppath)
  (let ((std-runname (conc "sched"  (time->string (seconds->local-time (current-seconds)) "%M%H%d"))))
    (with-queue-db
    (common:with-queue-db
     mtconf
     (lambda (pktsdirs pktsdir pdb)
       (let* ((rgconfdat (find-and-read-config (conc toppath "/runconfigs.config")))
	      (rgconf    (car rgconfdat))
	      (all-areas (map car (configf:get-section mtconf "areas")))
	      (contours  (configf:get-section mtconf "contours"))
	      (torun     (make-hash-table)) ;; target => ( ... info ... )

			(areas      (val-alist->areas    val-alist)) ;; areas can be a single string (a reference to call an areas function), or a list of area names.
			(dbdest     (alist-ref 'dbdest   val-alist))
			(appendconf (alist-ref 'appendconf val-alist))
			(file-globs (alist-ref 'glob val-alist))
			
			(runstarts  (find-pkts pdb '(runstart) `((o . ,contour)
								 (t . ,runkey))))
			(rspkts     (get-pkt-alists runstarts))
			(rspkts     (common:get-pkt-alists runstarts))
			;; starttimes is for run start times and is used to know when the last run was launched
			(starttimes (get-pkt-times rspkts)) ;; sort by age (youngest first) and delete duplicates by target
			(starttimes (common:get-pkt-times rspkts)) ;; sort by age (youngest first) and delete duplicates by target
			(last-run   (if (null? starttimes) ;; if '() then it has never been run, else get the max
					0
					(apply max (map cdr starttimes))))
			;; synctimes is for figuring out the last time a sync was done
			(syncstarts (find-pkts pdb '(syncstart) '())) ;; no qualifiers, a sync does all tarets etc.
			(sspkts     (get-pkt-alists syncstarts))
			(synctimes  (get-pkt-times  sspkts))
			(sspkts       (common:get-pkt-alists syncstarts))
			(synctimes    (common:get-pkt-times  sspkts))
			(last-sync  (if (null? synctimes) ;; if '() then it has never been run, else get the max
					0
					(apply max (map cdr synctimes))))
			)

		   (let ((delta (lambda (x)
				  (round (/ (- (current-seconds) x) 60)))))

		     exn
		     #f
		   (create-directory "logs")
		   #t)
		 #t)
	     "logs"
	     "/tmp")))
    (with-queue-db
    (common:with-queue-db
     mtconf
     (lambda (pktsdirs pktsdir pdb)
       (let* ((rgconfdat (find-and-read-config (conc toppath "/runconfigs.config")))
	      (rgconf    (car rgconfdat))
	      (areas     (configf:get-section mtconf "areas"))
	      (contours  (configf:get-section mtconf "contours"))
	      (pkts      (find-pkts pdb '(cmd) '()))

	 ;;  (hash-table-keys adjargs))
	 (let-values (((uuid pkt)
		       (command-line->pkt *action* adjargs #f)))
	   (write-pkt pktsdir uuid pkt))))
      ((dispatch import rungen process)
       (let* ((mtconfdat (simple-setup (args:get-arg "-start-dir")))
	      (mtconf    (car mtconfdat))
	      (toppath   (configf:lookup mtconf "dyndat" "toppath")))
	      (toppath   (configf:lookup mtconf "scratchdat" "toppath")))
	 (case (string->symbol *action*)
	   ((process)  (begin
			 (load-pkts-to-db mtconf)
			 (common:load-pkts-to-db mtconf)
			 (generate-run-pkts mtconf toppath)
			 (load-pkts-to-db mtconf)
			 (common:load-pkts-to-db mtconf)
			 (dispatch-commands mtconf toppath)))
	   ((import)   (load-pkts-to-db mtconf)) ;; import pkts
	   ((import)   (common:load-pkts-to-db mtconf)) ;; import pkts
	   ((rungen)   (generate-run-pkts mtconf toppath))
	   ((dispatch) (dispatch-commands mtconf toppath)))))
      ;; misc
      ((show)
       (if (> (length remargs) 0)
	   (let* ((mtconfdat (simple-setup (args:get-arg "-start-dir")))
		  (mtconf    (car mtconfdat))

    (iup:callback-set! *tim*
		       "ACTION_CB"
		       (lambda (x)
			 ;; Want to dedicate no more than 50% of the time to this so skip if
			 ;; 2x delta time has not passed since last query
			 (if (< nextmintime (current-milliseconds))
			     (let* ((starttime (current-milliseconds))
				    (changes   (dcommon:run-update keys data runname keypatts testpatt states statuses 'full my-window-id))
				    ;; (changes   (dcommon:run-update keys data runname keypatts testpatt states statuses 'full my-window-id))
				    (endtime   (current-milliseconds)))
			       (set! nextmintime (+ endtime (* 2 (- endtime starttime))))
			       (debug:print 11 *default-log-port* "CHANGE(S): " (car changes) "..."))
			       ;; (debug:print 11 *default-log-port* "CHANGE(S): " (car changes) "..."))
                               )
			     (debug:print-info 11 *default-log-port* "Server overloaded"))))))

;; (dboard:data-updaters-set! *data* (make-hash-table))
(newdashboard #f) ;; *dbstruct-local*)    
(iup:main-loop)

	 (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 0 *default-log-port* "exn=" (condition->list exn))
       (debug:print 5 *default-log-port* "exn=" (condition->list exn))
       (if (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)

(define (portlogger:get-prev-used-port db)
  (handle-exceptions
      exn
      (begin
	(debug:print 0 *default-log-port* "EXCEPTION: portlogger database probably overloaded or unreadable. If you see this message again remove /tmp/.$USER-portlogger.db")
	(debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	(debug:print 0 *default-log-port* "exn=" (condition->list exn))
	(debug:print 5 *default-log-port* "exn=" (condition->list exn))
	(print-call-chain (current-error-port))
	(debug:print 0 *default-log-port* "Continuing anyway.")
	#f)
    (sqlite3:fold-row
     (lambda (var curr)
       (or curr var curr))
     #f

		      (+ lowport ;; top of registered ports is 49152 but lets use ports in the registered range
			 (random (- 64000 lowport))))))
    (handle-exceptions
     exn
     (begin
       (debug:print 0 *default-log-port* "EXCEPTION: portlogger database probably overloaded or unreadable. If you see this message again remove /tmp/.$USER-portlogger.db")
       (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
       (debug:print 0 *default-log-port* "exn=" (condition->list exn))
       (debug:print 5 *default-log-port* "exn=" (condition->list exn))
       (print-call-chain (current-error-port))
       (debug:print 0 *default-log-port* "Continuing anyway."))
     (portlogger:take-port db portnum))
    portnum))

;; set port to "released", "failed" etc.
;; 

	 (numargs (length args))
	 (result  
	  (handle-exceptions
	   exn
	   (begin
	     (debug:print 0 *default-log-port* "EXCEPTION: portlogger database at " dbfname " probably overloaded or unreadable. Try removing it.")
	     (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	     (print "exn=" (condition->list exn))
	     (debug:print 5 *default-log-port* "exn=" (condition->list exn))
	     (debug:print 0 *default-log-port* " status:  " ((condition-property-accessor 'sqlite3 'status) exn))
	     (print-call-chain (current-error-port))
	     #f)
	   (case (string->symbol (car args)) ;; commands with two or more params
	     ((take)(portlogger:take-port db (string->number (cadr args))))
	     ((find)(portlogger:find-port db))
	     ((set) (let ((port  (cadr  args))

(define (process:cmd-run-proc-each-line cmd proc . params)
  ;; (print "Called with cmd=" cmd ", proc=" proc ", params=" params)
  (handle-exceptions
   exn
   (begin
     (print "ERROR:  Failed to run command: " cmd " " (string-intersperse params " "))
     (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
     (print "exn=" (condition->list exn))
     (debug:print 5 *default-log-port* "exn=" (condition->list exn))
     #f)
   (let-values (((fh fho pid) (if (null? params)
				  (process cmd)
				  (process cmd params))))
       (let loop ((curr (read-line fh))
		(result  '()))
       (if (not (eof-object? curr))

  (let* ((run-info             (rmt:get-run-info run-id))
	(tests-info            (mt:get-tests-for-run run-id #f '() '())) ;;  qryvals: "id,testname,item_path"))
	(sorted-test-names     (tests:sort-by-priority-and-waiton test-records))
	(test-registry         (make-hash-table))
	(registry-mutex        (make-mutex))
	(num-retries           0)
	(max-retries           (config-lookup *configdat* "setup" "maxretries"))
	(max-concurrent-jobs   (let ((mcj (config-lookup *configdat* "setup"     "max_concurrent_jobs")))
	(max-concurrent-jobs   (configf:lookup-number *configdat* "setup" "max_concurrent_jobs" default: 50))
				 (if (and mcj (string->number mcj))
				     (string->number mcj)
				     1))) ;; length of the register queue ahead
	(reglen                (if (number? reglen-in) reglen-in 1))
        (reglen                (if (number? reglen-in) reglen-in 1))
	(last-time-incomplete  (- (current-seconds) 900)) ;; force at least one clean up cycle
	(last-time-some-running (current-seconds))
	;; (tdbdat                (tasks:open-db))
	(runsdat (make-runs:dat
		  ;; hed: hed
		  ;; tal: tal
		  ;; reg: reg

;; Copyright 2006-2017, Matthew Welland.
;; 
;;  This program is made available under the GNU GPL version 2.0 or
;;  greater. See the accompanying file COPYING for details.
;; 
;;  This program is distributed WITHOUT ANY WARRANTY; without even the
;;  implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
;;  PURPOSE.

(require-extension (srfi 18) extras tcp s11n)

(use srfi-1 posix regex regex-case srfi-69 hostinfo md5 message-digest directory-utils posix-extras matchable)
(use srfi-1 posix regex regex-case srfi-69 hostinfo md5 message-digest directory-utils posix-extras matchable
     )

(use spiffy uri-common intarweb http-client spiffy-request-vars)

(declare (unit server))

(declare (uses common))
(declare (uses db))
(declare (uses tasks)) ;; tasks are where stuff is maintained about what is running.
(declare (uses synchash))
;; (declare (uses synchash))
(declare (uses http-transport))
(declare (uses launch))
(declare (uses daemon))

(include "common_records.scm")
(include "db_records.scm")

(define (server:make-server-url hostport)
  (if (not hostport)
      #f
      (conc "http://" (car hostport) ":" (cadr hostport))))

(define  *server-loop-heart-beat* (current-seconds))

;;======================================================================
;; P K T S   S T U F F 
;;======================================================================

;; ???

;;======================================================================
;; S E R V E R
;;======================================================================

;; Call this to start the actual server
;;

      (debug:print-error 0 *default-log-port* "Called tasks:wait-on-journal with path=" path " (not a string)")
      (let ((fullpath (conc path "-journal")))
	(handle-exceptions
	 exn
	 (begin
	   (print-call-chain (current-error-port))
	   (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	   (debug:print 0 *default-log-port* " exn=" (condition->list exn))
	   (debug:print 5 *default-log-port* " exn=" (condition->list exn))
	   (debug:print 0 *default-log-port* "tasks:wait-on-journal failed. Continuing on, you can ignore this call-chain")
	   #t) ;; if stuff goes wrong just allow it to move on
	 (let loop ((journal-exists (file-exists? fullpath))
		    (count          n)) ;; wait ten times ...
	   (if journal-exists
	       (begin
		 (if (and waiting-msg

      *task-db*
      (handle-exceptions
       exn
       (if (> numretries 0)
	   (begin
	     (print-call-chain (current-error-port))
	     (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	     (debug:print 0 *default-log-port* " exn=" (condition->list exn))
	     (debug:print 5 *default-log-port* " exn=" (condition->list exn))
	     (thread-sleep! 1)
	     (tasks:open-db numretries (- numretries 1)))
	   (begin
	     (print-call-chain (current-error-port))
	     (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	     (debug:print 0 *default-log-port* " exn=" (condition->list exn))))
	     (debug:print 5 *default-log-port* " exn=" (condition->list exn))))
       (let* ((dbpath        (db:dbfile-path )) ;; (tasks:get-task-db-path))
	      (dbfile       (conc dbpath "/monitor.db"))
	      (avail        (tasks:wait-on-journal dbpath 10)) ;; wait up to about 10 seconds for the journal to go away
	      (exists       (file-exists? dbpath))
	      (write-access (file-write-access? dbpath))
	      (mdb          (cond ;; what the hek is *toppath* doing here?
			     ((and (string? *toppath*)(file-write-access? *toppath*))

  (equal? (configf:lookup *configdat* "server" "required") "yes"))

;; no elegance here ...
;;
(define (tasks:kill-server hostname pid #!key (kill-switch ""))
  (debug:print-info 0 *default-log-port* "Attempting to kill server process " pid " on host " hostname)
  (setenv "TARGETHOST" hostname)
  (let ((logdir (if (directory-exists? "logs")
  (let* ((logdir (if (directory-exists? "logs")
                    "logs/"
                    "")))
                    ""))
         (logfile (if logdir (conc "logs/server-"pid"-"hostname".log") #f))
         (gzfile  (if logfile (conc logfile ".gz"))))
    (setenv "TARGETHOST_LOGF" (conc logdir "server-kills.log"))

    (system (conc "nbfake kill "kill-switch" "pid))
    
    (unsetenv "TARGETHOST_LOGF")
    (unsetenv "TARGETHOST")))

    (when logfile
      (thread-sleep! 0.5)
      (if (file-exists? gzfile) (delete-file gzfile))
      (system (conc "gzip "logfile))
      
      (unsetenv "TARGETHOST_LOGF")
      (unsetenv "TARGETHOST"))))
    
 
;;======================================================================
;; M O N I T O R S
;;======================================================================

(define (tasks:remove-monitor-record mdb)
  (sqlite3:execute mdb "DELETE FROM monitors WHERE pid=? AND hostname=?;"

	   (set! remtries (- remtries 1))
	   (thread-sleep! 10)
	   (tests:set-full-meta-info db test-id run-id minutes work-area (- remtries 1)))
	 (let ((err-status ((condition-property-accessor 'sqlite3 'status #f) exn)))
	   (debug:print-error 0 *default-log-port* "tried for over a minute to update meta info and failed. Giving up")
	   (debug:print 0 *default-log-port* "EXCEPTION: database probably overloaded or unreadable.")
	   (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))
	   (print "exn=" (condition->list exn))
	   (debug:print 5 *default-log-port* "exn=" (condition->list exn))
	   (debug:print 0 *default-log-port* " status:  " ((condition-property-accessor 'sqlite3 'status) exn))
	   (print-call-chain (current-error-port))))
     (tests:update-testdat-meta-info db test-id work-area cpuload diskfree minutes)
  )))
	 
;;======================================================================
;; A R C H I V I N G

# this may save a few milliseconds on launching tests
# launchwait no
waivercommentpatt ^WW\d+ [a-z].*
incomplete-timeout 1

# wait 0.5 seconds between launching every process
#
launch-delay 0.5
# launch-delay 0.5
launch-delay 0


# wait for runs to completely complete. yes, anything else is no
run-wait yes

# If set to "default" the old code is used. Otherwise defaults to 200 or uses
# numeric value given.
#

(declare (unit tree))
(declare (uses margs))
(declare (uses launch))
(declare (uses megatest-version))
(declare (uses gutils))
(declare (uses db))
(declare (uses server))
(declare (uses synchash))
;; (declare (uses synchash))
(declare (uses dcommon))

(include "common_records.scm")
(include "db_records.scm")
(include "key_records.scm")

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