Overview
Comment: | Added journal based statical droop based throttling of queries. |
---|---|
Downloads: | Tarball | ZIP archive | SQL archive |
Timelines: | family | ancestors | descendants | both | v1.81-journal-based-throttling |
Files: | files | file ages | folders |
SHA1: |
fc6b05f92450cc3eeea48602396e7d47 |
User & Date: | mrwellan on 2024-07-10 17:44:52 |
Other Links: | branch diff | manifest | tags |
Context
2024-07-10
| ||
18:10 | Gate test launch based on journal load. Values from load calc seem wrong. Should be 0-1.0 but seeing integers 0, 1, 2 ... check-in: 2635b582e7 user: mrwellan tags: v1.81-journal-based-throttling | |
17:44 | Added journal based statical droop based throttling of queries. check-in: fc6b05f924 user: mrwellan tags: v1.81-journal-based-throttling | |
09:17 | Added docs/csirc which works with both chicken 4 and 5 check-in: e75a04de3e user: mrwellan tags: v1.81-journal-based-throttling | |
Changes
Modified rmt.scm from [519878889b] to [a86edf18db].
︙ | ︙ | |||
90 91 92 93 94 95 96 | ;; NB// area-dat replaced by ttdat ;; (define (rmt:send-receive cmd run-id params #!key (attemptnum 1)(ttdat #f)) (assert (or (not run-id) (number? run-id)) "FATAL: run-id is required to be a number or #f") (assert *toppath* "FATAL: rmt:send-receive called with *toppath* not set.") (let* ((areapath *toppath*) ;; TODO - resolve from dbstruct to be compatible with multiple areas (readonly-mode (rmtmod:calc-ro-mode ttdat *toppath*)) | | > > > > > > > > > > > > > > > < | 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 | ;; NB// area-dat replaced by ttdat ;; (define (rmt:send-receive cmd run-id params #!key (attemptnum 1)(ttdat #f)) (assert (or (not run-id) (number? run-id)) "FATAL: run-id is required to be a number or #f") (assert *toppath* "FATAL: rmt:send-receive called with *toppath* not set.") (let* ((areapath *toppath*) ;; TODO - resolve from dbstruct to be compatible with multiple areas (readonly-mode (rmtmod:calc-ro-mode ttdat *toppath*)) (testsuite (common:get-testsuite-name)) (dbfname (conc (dbfile:run-id->dbnum run-id)".db")) (dbdir (conc areapath "/.mtdb"))) (if (and (not *journal-stats*) (file-exists? dbdir)) (tt:start-stats dbdir)) ;; fixme - find the right call to get the db directory ;; check the load on dbfname and add some delay using a droop curve of sorts (if *journal-stats* (let* ((stats (tt:get-journal-stats)) (load (or (alist-ref dbfname stats equal?) 0))) (if (> load 0) (let ((dely (* 10 load))) (debug:print 0 *default-log-port* "Journal load "load" delaying queries "dely"s.") (thread-sleep! dely))))) (case (rmt:transport-mode) ((tcp) (let* ((start-time (current-seconds)) ;; snapshot time so all use cases get same value (attemptnum (+ 1 attemptnum)) (mtexe (common:find-local-megatest)) (ttdat (rmt:set-ttdat areapath ttdat)) (conn (tt:get-conn ttdat dbfname)) (is-main (equal? dbfname "main.db")) ;; why not (not run-id) ? (server-start-proc (if is-main #f (lambda () ;; (debug:print-info 0 *default-log-port* "starting server for dbfname: "dbfname) |
︙ | ︙ |
Modified tcp-transportmod.scm from [2f9da6d8df] to [98a778bd3e].
︙ | ︙ | |||
1138 1139 1140 1141 1142 1143 1144 | (defstruct jstats (count 0) (jcount (make-hash-table)) ;; 1.db => journal_count ) ;; timeblk => jstats | | | | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 | (defstruct jstats (count 0) (jcount (make-hash-table)) ;; 1.db => journal_count ) ;; timeblk => jstats (define *journal-stats* #f) ;; (make-hash-table)) ;; monte-carlo-esque random sampling of journal files ;; for all the files: ;; if .journal ;; update stats +1 +1 ;; update stats +1 0 ;; (define (tt:write-load-tracking dbdir) (let* ((cs (current-seconds)) (key (inexact->exact (quotient cs 10))) (old (- key 5)) ;; 4 x 10 seconds ago (jstat (if (hash-table-exists? *journal-stats* key) (hash-table-ref *journal-stats* key ) (let ((new (make-jstats))) (hash-table-set! *journal-stats* key new) new)))) ;; clear out old records (for-each (lambda (key) (if (< key old) (hash-table-delete! *journal-stats* key))) (hash-table-keys *journal-stats*)) ;; increment our count of observations (jstats-count-set! jstat (+ (jstats-count jstat) 1)) ;; now find and increment journal file counts (directory-fold (lambda (fname res) ;; is it a journal file? (let ((parts (string-match "^(.*\\.db)-journal.*" fname))) (match parts ((_ dbfname) (hash-table-set! (jstats-jcount jstat) dbfname (+ (hash-table-ref/default (jstats-jcount jstat) dbfname 0) 1) )) (else #f) ))) '() dbdir ))) (define *journal-stats-mutex* (make-mutex)) (define (tt:journal-stats-run dbdir) (if (not *journal-stats*)(set! *journal-stats* (make-hash-table))) (let loop () (mutex-lock! *journal-stats-mutex*) (tt:write-load-tracking dbdir) (mutex-unlock! *journal-stats-mutex*) (thread-sleep! (/ (random 1000) 100.0)) (loop))) ;; call this to start a thread that is keeping the journal-stats up to date. (define (tt:start-stats dbdir) (thread-start! (make-thread (lambda ()(tt:journal-stats-run dbdir)) "Journal stats collection thread"))) (define (tt:get-journal-stats) (let* ((result (make-jstats)) (hitcounts (jstats-jcount result))) (if *journal-stats* (begin (mutex-lock! *journal-stats-mutex*) (hash-table-for-each *journal-stats* (lambda (k v) ;; key jstats (let* ((count (jstats-count v)) (jcount (jstats-jcount v))) ;; dbfname => hit count (jstats-count-set! result (+ (jstats-count result) (jstats-count v))) (hash-table-for-each jcount (lambda (dbfname hit-count) (hash-table-set! hitcounts dbfname (+ hit-count (hash-table-ref/default hitcounts dbfname 0)))))))) (mutex-unlock! *journal-stats-mutex*)) (debug:print 0 *default-log-port* "INFO: *journal-stats* not set.")) ;; convert to normalized alist (let ((tot (min (jstats-count result) 1)) ;; avoid divide by zero (hits (jstats-jcount result))) ;; 1.db => count (hash-table-map hits (lambda (fname hitcount) (cons fname (/ hitcount tot))))) )) ;; megatest> (import tcp-transportmod) ;; megatest> (tt:write-load-tracking ".mtdb") ;; megatest> (hash-table-keys *journal-stats*) ;; (172060297) ;; megatest> (jstats->alist (hash-table-ref *journal-stats* 172060297)) ;; ((count . 1) (jcount . #<hash-table (1)>)) ;; megatest> (jstats-jcount (hash-table-ref *journal-stats* 172060297)) ;; #<hash-table (1)> ;; megatest> (hash-table->alist (jstats-jcount (hash-table-ref *journal-stats* 172060297))) ;; (("1.db" . 4)) ) |