Megatest

Check-in [48bb95d322]
Login
Overview
Comment:removed -daemonize when starting server; removed local fallback in rmt:send-receive
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | rpc-transport
Files: files | file ages | folders
SHA1: 48bb95d322d9c5fc5e3d08af31815e8ad9907934
User & Date: bjbarcla on 2016-11-14 15:58:48
Other Links: branch diff | manifest | tags
Context
2016-11-14
17:55
caught another unimplemented area. check-in: a827c0e1f8 user: bjbarcla tags: rpc-transport
15:58
removed -daemonize when starting server; removed local fallback in rmt:send-receive check-in: 48bb95d322 user: bjbarcla tags: rpc-transport
15:47
removed -daemonize when starting server; removed local fallback in rmt:send-receive check-in: a31c1d5781 user: bjbarcla tags: rpc-transport
Changes

Modified rmt.scm from [a521022e85] to [a1a9e49fdf].

86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151

(define *send-receive-mutex* (make-mutex)) ;; should have separate mutex per run-id

;; RA => e.g. usage (rmt:send-receive 'get-var #f (list varname))
;;
(define (rmt:send-receive cmd rid params #!key (attemptnum 1)) ;; start attemptnum at 1 so the modulo below works as expected
  ;; side-effect: clean out old connections
  ;; (mutex-lock! *db-multi-sync-mutex*)
  (let ((expire-time (- (current-seconds) (server:get-timeout) 10))) ;; don't forget the 10 second margin
    (for-each 
     (lambda (run-id)
       (let ((connection (hash-table-ref/default *runremote* run-id #f)))
         (if (and (vector? connection)
        	  (< (http-transport:server-dat-get-last-access connection) expire-time)) ;; BB> BBTODO: make this generic, not http transport specific.
             (begin
               (debug:print-info 0 *default-log-port* "Discarding connection to server for run-id " run-id ", too long between accesses")
               (hash-table-delete! *runremote* run-id)))))
     (hash-table-keys *runremote*)))
  
  ;; (mutex-unlock! *db-multi-sync-mutex*)
  ;; (mutex-lock! *send-receive-mutex*)
  (let* ((run-id          (if rid rid 0))
	 (connection-info (rmt:get-connection-info run-id)))
    ;; the nmsg method does the encoding under the hood (the http method should be changed to do this also)
    (if connection-info
	;; use the server if have connection info
	(let* ((transport-type (rmt:run-id->transport-type run-id))

               ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
               ;;  Here, we make request to remote server
               ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
               (dat     (case transport-type ;; BB: replaced *transport-type* global with run-id specific transport-type
			  ((http)(condition-case
				  (http-transport:client-api-send-receive run-id connection-info cmd params)
				  ((commfail)(vector #f "communications fail"))
				  ((exn)(vector #f "other fail"))))
                          ((rpc) (rpc-transport:client-api-send-receive run-id connection-info cmd params)) ;; BB: let us error out for now
			  (else  
                           (debug:print-error 0 *default-log-port* "(1) Transport [" transport-type
                                              "] specified for run-id [" run-id
                                              "] is not implemented in rmt:send-receive.  Cannot proceed." (symbol? transport-type))
                           (vector #f (conc "transport ["transport-type"] unimplemented")))))

               
	       (success (if (vector? dat) (vector-ref dat 0) #f))
	       (res     (if (vector? dat) (vector-ref dat 1) #f)))
	  (if (vector? connection-info)(http-transport:server-dat-update-last-access connection-info)) ;; BB> BBTODO: make this generic, not http transport specific.
	  (if success
	      (begin
		;; (mutex-unlock! *send-receive-mutex*)
		(case transport-type 
		  ((http rpc) res) ;; (db:string->obj res))
                  (else
                   (debug:print-error 0 *default-log-port* "(2) Transport [" transport-type
                                      "] specified for run-id [" run-id
                                      "] is not implemented in rmt:send-receive.  Cannot proceed. Also unexpected since this branch follows success which would follow a suported transport...")
                   #f)
                  )) ;; (vector-ref res 1)))

              ;; no success...
	      (begin ;; let ((new-connection-info (client:setup run-id)))
		(debug:print 0 *default-log-port* "WARNING: Communication failed, trying call to rmt:send-receive again.")
                (case transport-type
                  
                  ((http rpc)
                   (hash-table-delete! *runremote* run-id) ;; don't keep using the same connection







<











<
<
|





|














|















|







86
87
88
89
90
91
92

93
94
95
96
97
98
99
100
101
102
103


104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148

(define *send-receive-mutex* (make-mutex)) ;; should have separate mutex per run-id

;; RA => e.g. usage (rmt:send-receive 'get-var #f (list varname))
;;
(define (rmt:send-receive cmd rid params #!key (attemptnum 1)) ;; start attemptnum at 1 so the modulo below works as expected
  ;; side-effect: clean out old connections

  (let ((expire-time (- (current-seconds) (server:get-timeout) 10))) ;; don't forget the 10 second margin
    (for-each 
     (lambda (run-id)
       (let ((connection (hash-table-ref/default *runremote* run-id #f)))
         (if (and (vector? connection)
        	  (< (http-transport:server-dat-get-last-access connection) expire-time)) ;; BB> BBTODO: make this generic, not http transport specific.
             (begin
               (debug:print-info 0 *default-log-port* "Discarding connection to server for run-id " run-id ", too long between accesses")
               (hash-table-delete! *runremote* run-id)))))
     (hash-table-keys *runremote*)))
  


  (let* ((run-id     (if rid rid 0))
	 (connection-info (rmt:get-connection-info run-id)))
    ;; the nmsg method does the encoding under the hood (the http method should be changed to do this also)
    (if connection-info
	;; use the server if have connection info
	(let* ((transport-type (rmt:run-id->transport-type run-id))
               
               ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
               ;;  Here, we make request to remote server
               ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
               (dat     (case transport-type ;; BB: replaced *transport-type* global with run-id specific transport-type
			  ((http)(condition-case
				  (http-transport:client-api-send-receive run-id connection-info cmd params)
				  ((commfail)(vector #f "communications fail"))
				  ((exn)(vector #f "other fail"))))
                          ((rpc) (rpc-transport:client-api-send-receive run-id connection-info cmd params)) ;; BB: let us error out for now
			  (else  
                           (debug:print-error 0 *default-log-port* "(1) Transport [" transport-type
                                              "] specified for run-id [" run-id
                                              "] is not implemented in rmt:send-receive.  Cannot proceed." (symbol? transport-type))
                           (vector #f (conc "transport ["transport-type"] unimplemented")))))
               
               
	       (success (if (vector? dat) (vector-ref dat 0) #f))
	       (res     (if (vector? dat) (vector-ref dat 1) #f)))
	  (if (vector? connection-info)(http-transport:server-dat-update-last-access connection-info)) ;; BB> BBTODO: make this generic, not http transport specific.
	  (if success
	      (begin
		;; (mutex-unlock! *send-receive-mutex*)
		(case transport-type 
		  ((http rpc) res) ;; (db:string->obj res))
                  (else
                   (debug:print-error 0 *default-log-port* "(2) Transport [" transport-type
                                      "] specified for run-id [" run-id
                                      "] is not implemented in rmt:send-receive.  Cannot proceed. Also unexpected since this branch follows success which would follow a suported transport...")
                   #f)
                  )) ;; (vector-ref res 1)))
              
              ;; no success...
	      (begin ;; let ((new-connection-info (client:setup run-id)))
		(debug:print 0 *default-log-port* "WARNING: Communication failed, trying call to rmt:send-receive again.")
                (case transport-type
                  
                  ((http rpc)
                   (hash-table-delete! *runremote* run-id) ;; don't keep using the same connection
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
                   ;; (thread-sleep! 2)
                   (rmt:send-receive cmd run-id params attemptnum: (+ attemptnum 1)))
                  (else
                   (debug:print-error 0 *default-log-port* "(3) Transport [" transport-type
                                      "] specified for run-id [" run-id
                                      "] is not implemented in rmt:send-receive.  Cannot proceed.")
                   #f)))))

	;; no connection info; try to start a server
	;;
	;; Note: The tasks db was checked for a server in starting mode in the rmt:get-connection-info call
	;;
        (let* ((faststart (configf:lookup *configdat* "server" "faststart")))
          (hash-table-delete! *runremote* run-id)
          ;; (mutex-unlock! *send-receive-mutex*)
          (if (and faststart (equal? faststart "no"))
              (begin
                (tasks:start-and-wait-for-server (db:delay-if-busy (tasks:open-db)) run-id 10)
                (thread-sleep! (random 5)) ;; give some time to settle and minimize collison?
                (rmt:send-receive cmd rid params attemptnum: (+ attemptnum 1)))
              (let ((start-time (current-milliseconds))
                    (max-query  (string->number (or (configf:lookup *configdat* "server" "server-query-threshold")
                                                    "300")))
                    (newres     (rmt:open-qry-close-locally cmd run-id params)))
                (let ((delta (- (current-milliseconds) start-time)))
                  (if (> delta max-query)
                      (begin
                        (debug:print-info 0 *default-log-port* "Starting server as query time " delta " is over the limit of " max-query)
                        (server:kind-run run-id)))
                  ;; return the result!
                  newres)))))))

(define (rmt:update-db-stats run-id rawcmd params duration)
  (mutex-lock! *db-stats-mutex*)
  (handle-exceptions
   exn
   (begin
     (debug:print 0 *default-log-port* "WARNING: stats collection failed in update-db-stats")







|




<
<
<
<
|
|
|
|
|
<
<
<
<
<
<
<
<
<
<







160
161
162
163
164
165
166
167
168
169
170
171




172
173
174
175
176










177
178
179
180
181
182
183
                   ;; (thread-sleep! 2)
                   (rmt:send-receive cmd run-id params attemptnum: (+ attemptnum 1)))
                  (else
                   (debug:print-error 0 *default-log-port* "(3) Transport [" transport-type
                                      "] specified for run-id [" run-id
                                      "] is not implemented in rmt:send-receive.  Cannot proceed.")
                   #f)))))
        
	;; no connection info; try to start a server
	;;
	;; Note: The tasks db was checked for a server in starting mode in the rmt:get-connection-info call
	;;




        (begin
          (tasks:start-and-wait-for-server (db:delay-if-busy (tasks:open-db)) run-id 10)
          (thread-sleep! (random 5)) ;; give some time to settle and minimize collison?
          (rmt:send-receive cmd rid params attemptnum: (+ attemptnum 1))))))
  











(define (rmt:update-db-stats run-id rawcmd params duration)
  (mutex-lock! *db-stats-mutex*)
  (handle-exceptions
   exn
   (begin
     (debug:print 0 *default-log-port* "WARNING: stats collection failed in update-db-stats")
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
			     (if (> tot 10)
				 (cons newmax-cmd currmax)
				 (cons 'none 0))
			     (loop (car tal)(cdr tal) newmax-cmd currmax)))))))
    (mutex-unlock! *db-stats-mutex*)
    res))
	  
(define (rmt:open-qry-close-locally cmd run-id params #!key (remretries 5))
  (let* ((dbstruct-local (if *dbstruct-db*
			     *dbstruct-db*
			     (let* ((dbdir (db:dbfile-path #f)) ;;  (conc    (configf:lookup *configdat* "setup" "linktree") "/.db"))
				    (db (make-dbr:dbstruct path:  dbdir local: #t)))
			       (set! *dbstruct-db* db)
			       db)))
	 (db-file-path   (db:dbfile-path 0))
	 ;; (read-only      (not (file-read-access? db-file-path)))
	 (start          (current-milliseconds))
	 (resdat         (api:execute-requests dbstruct-local (vector (symbol->string cmd) params)))
	 (success        (vector-ref resdat 0))
	 (res            (vector-ref resdat 1))
	 (duration       (- (current-milliseconds) start)))
    (if (not success)
	(if (> remretries 0)
	    (begin
	      (debug:print-error 0 *default-log-port* "local query failed. Trying again.")
	      (thread-sleep! (/ (random 5000) 1000)) ;; some random delay 
	      (rmt:open-qry-close-locally cmd run-id params remretries: (- remretries 1)))
	    (begin
	      (debug:print-error 0 *default-log-port* "too many retries in rmt:open-qry-close-locally, giving up")
	      #f))
	(begin
	  ;; (rmt:update-db-stats run-id cmd params duration)
	  ;; mark this run as dirty if this was a write
	  (if (not (member cmd api:read-only-queries))
	      (let ((start-time (current-seconds)))
		(mutex-lock! *db-multi-sync-mutex*)
		;; (if (not (hash-table-ref/default *db-local-sync* run-id #f))
		;; just set it every time. Is a write more expensive than a read and does it matter?
		(hash-table-set! *db-local-sync* (or run-id 0) start-time) ;; the oldest "write"
		(mutex-unlock! *db-multi-sync-mutex*)))
	  res))))

(define (rmt:send-receive-no-auto-client-setup connection-info cmd run-id params)
  (let* ((run-id   (if run-id run-id 0))
	 ;; (jparams  (db:obj->string params)) ;; (rmt:dat->json-str params))
	 (res  	   (handle-exceptions
		    exn
		    #f
		    (http-transport:client-api-send-receive run-id connection-info cmd params))))







<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<







228
229
230
231
232
233
234



































235
236
237
238
239
240
241
			     (if (> tot 10)
				 (cons newmax-cmd currmax)
				 (cons 'none 0))
			     (loop (car tal)(cdr tal) newmax-cmd currmax)))))))
    (mutex-unlock! *db-stats-mutex*)
    res))
	  



































(define (rmt:send-receive-no-auto-client-setup connection-info cmd run-id params)
  (let* ((run-id   (if run-id run-id 0))
	 ;; (jparams  (db:obj->string params)) ;; (rmt:dat->json-str params))
	 (res  	   (handle-exceptions
		    exn
		    #f
		    (http-transport:client-api-send-receive run-id connection-info cmd params))))