maintenance/runJobs.php

   1 <?php
   2 /**
   3  * Run pending jobs.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Maintenance
  22  */
  23
  24 require_once __DIR__ . '/Maintenance.php';
  25
  26 /**
  27  * Maintenance script that runs pending jobs.
  28  *
  29  * @ingroup Maintenance
  30  */
  31 class RunJobs extends Maintenance {
  32         public function __construct() {
  33                 parent::__construct();
  34                 $this->mDescription = "Run pending jobs";
  35                 $this->addOption( 'maxjobs', 'Maximum number of jobs to run', false, true );
  36                 $this->addOption( 'maxtime', 'Maximum amount of wall-clock time', false, true );
  37                 $this->addOption( 'type', 'Type of job to run', false, true );
  38                 $this->addOption( 'procs', 'Number of processes to use', false, true );
  39                 $this->addOption( 'nothrottle', 'Ignore job throttling configuration', false, false );
  40         }
  41
  42         public function memoryLimit() {
  43                 if ( $this->hasOption( 'memory-limit' ) ) {
  44                         return parent::memoryLimit();
  45                 }
  46
  47                 // Don't eat all memory on the machine if we get a bad job.
  48                 return "150M";
  49         }
  50
  51         public function execute() {
  52                 if ( wfReadOnly() ) {
  53                         $this->error( "Unable to run jobs; the wiki is in read-only mode.", 1 ); // die
  54                 }
  55
  56                 if ( $this->hasOption( 'procs' ) ) {
  57                         $procs = intval( $this->getOption( 'procs' ) );
  58                         if ( $procs < 1 || $procs > 1000 ) {
  59                                 $this->error( "Invalid argument to --procs", true );
  60                         } elseif ( $procs != 1 ) {
  61                                 $fc = new ForkController( $procs );
  62                                 if ( $fc->start() != 'child' ) {
  63                                         exit( 0 );
  64                                 }
  65                         }
  66                 }
  67
  68                 $type = $this->getOption( 'type', false );
  69                 $maxJobs = $this->getOption( 'maxjobs', false );
  70                 $maxTime = $this->getOption( 'maxtime', false );
  71                 $noThrottle = $this->hasOption( 'nothrottle' );
  72                 $startTime = time();
  73
  74                 $group = JobQueueGroup::singleton();
  75                 // Handle any required periodic queue maintenance
  76                 $count = $group->executeReadyPeriodicTasks();
  77                 if ( $count > 0 ) {
  78                         $this->runJobsLog( "Executed $count periodic queue task(s)." );
  79                 }
  80
  81                 $backoffs = $this->loadBackoffs(); // map of (type => UNIX expiry)
  82                 $startingBackoffs = $backoffs; // avoid unnecessary writes
  83                 $backoffExpireFunc = function ( $t ) {
  84                         return $t > time();
  85                 };
  86
  87                 $jobsRun = 0; // counter
  88                 $flags = JobQueueGroup::USE_CACHE;
  89                 $lastTime = time(); // time since last slave check
  90                 do {
  91                         $backoffs = array_filter( $backoffs, $backoffExpireFunc );
  92                         $blacklist = $noThrottle ? array() : array_keys( $backoffs );
  93                         if ( $type === false ) {
  94                                 $job = $group->pop( JobQueueGroup::TYPE_DEFAULT, $flags, $blacklist );
  95                         } elseif ( in_array( $type, $blacklist ) ) {
  96                                 $job = false; // requested queue in backoff state
  97                         } else {
  98                                 $job = $group->pop( $type ); // job from a single queue
  99                         }
 100                         if ( $job ) { // found a job
 101                                 ++$jobsRun;
 102                                 $this->runJobsLog( $job->toString() . " STARTING" );
 103
 104                                 // Set timer to stop the job if too much CPU time is used
 105                                 set_time_limit( $maxTime ? : 0 );
 106                                 // Run the job...
 107                                 wfProfileIn( __METHOD__ . '-' . get_class( $job ) );
 108                                 $t = microtime( true );
 109                                 try {
 110                                         $status = $job->run();
 111                                         $error = $job->getLastError();
 112                                 } catch ( MWException $e ) {
 113                                         MWExceptionHandler::rollbackMasterChangesAndLog( $e );
 114                                         $status = false;
 115                                         $error = get_class( $e ) . ': ' . $e->getMessage();
 116                                         $e->report(); // write error to STDERR and the log
 117                                 }
 118                                 $timeMs = intval( ( microtime( true ) - $t ) * 1000 );
 119                                 wfProfileOut( __METHOD__ . '-' . get_class( $job ) );
 120                                 // Disable the timer
 121                                 set_time_limit( 0 );
 122
 123                                 // Mark the job as done on success or when the job cannot be retried
 124                                 if ( $status !== false || !$job->allowRetries() ) {
 125                                         $group->ack( $job ); // done
 126                                 }
 127
 128                                 if ( $status === false ) {
 129                                         $this->runJobsLog( $job->toString() . " t=$timeMs error={$error}" );
 130                                 } else {
 131                                         $this->runJobsLog( $job->toString() . " t=$timeMs good" );
 132                                 }
 133
 134                                 // Back off of certain jobs for a while (for throttling and for errors)
 135                                 $ttw = $this->getBackoffTimeToWait( $job );
 136                                 if ( $status === false && mt_rand( 0, 49 ) == 0 ) {
 137                                         $ttw = max( $ttw, 30 );
 138                                 }
 139                                 if ( $ttw > 0 ) {
 140                                         $jType = $job->getType();
 141                                         $backoffs[$jType] = isset( $backoffs[$jType] ) ? $backoffs[$jType] : 0;
 142                                         $backoffs[$jType] = max( $backoffs[$jType], time() + $ttw );
 143                                 }
 144
 145                                 // Break out if we hit the job count or wall time limits...
 146                                 if ( $maxJobs && $jobsRun >= $maxJobs ) {
 147                                         break;
 148                                 } elseif ( $maxTime && ( time() - $startTime ) > $maxTime ) {
 149                                         break;
 150                                 }
 151
 152                                 // Don't let any of the main DB slaves get backed up
 153                                 $timePassed = time() - $lastTime;
 154                                 if ( $timePassed >= 5 || $timePassed < 0 ) {
 155                                         wfWaitForSlaves();
 156                                         $lastTime = time();
 157                                 }
 158                                 // Don't let any queue slaves/backups fall behind
 159                                 if ( $jobsRun > 0 && ( $jobsRun % 100 ) == 0 ) {
 160                                         $group->waitForBackups();
 161                                 }
 162
 163                                 // Bail if near-OOM instead of in a job
 164                                 $this->assertMemoryOK();
 165                         }
 166                 } while ( $job ); // stop when there are no jobs
 167                 // Sync the persistent backoffs for the next runJobs.php pass
 168                 $backoffs = array_filter( $backoffs, $backoffExpireFunc );
 169                 if ( $backoffs !== $startingBackoffs ) {
 170                         $this->syncBackoffs( $backoffs );
 171                 }
 172         }
 173
 174         /**
 175          * @param Job $job
 176          * @return int Seconds for this runner to avoid doing more jobs of this type
 177          * @see $wgJobBackoffThrottling
 178          */
 179         private function getBackoffTimeToWait( Job $job ) {
 180                 global $wgJobBackoffThrottling;
 181
 182                 if ( !isset( $wgJobBackoffThrottling[$job->getType()] ) ||
 183                         $job instanceof DuplicateJob // no work was done
 184                 ) {
 185                         return 0; // not throttled
 186                 }
 187
 188                 $itemsPerSecond = $wgJobBackoffThrottling[$job->getType()];
 189                 if ( $itemsPerSecond <= 0 ) {
 190                         return 0; // not throttled
 191                 }
 192
 193                 $seconds = 0;
 194                 if ( $job->workItemCount() > 0 ) {
 195                         $exactSeconds = $job->workItemCount() / $itemsPerSecond;
 196                         // use randomized rounding
 197                         $seconds = floor( $exactSeconds );
 198                         $remainder = $exactSeconds - $seconds;
 199                         $seconds += ( mt_rand() / mt_getrandmax() < $remainder ) ? 1 : 0;
 200                 }
 201
 202                 return (int)$seconds;
 203         }
 204
 205         /**
 206          * Get the previous backoff expiries from persistent storage
 207          *
 208          * @return array Map of (job type => backoff expiry timestamp)
 209          */
 210         private function loadBackoffs() {
 211                 $section = new ProfileSection( __METHOD__ );
 212
 213                 $backoffs = array();
 214                 $file = wfTempDir() . '/mw-runJobs-backoffs.json';
 215                 if ( is_file( $file ) ) {
 216                         $handle = fopen( $file, 'rb' );
 217                         flock( $handle, LOCK_SH );
 218                         $content = stream_get_contents( $handle );
 219                         flock( $handle, LOCK_UN );
 220                         fclose( $handle );
 221                         $backoffs = json_decode( $content, true ) ? : array();
 222                 }
 223
 224                 return $backoffs;
 225         }
 226
 227         /**
 228          * Merge the current backoff expiries from persistent storage
 229          *
 230          * @param array $backoffs Map of (job type => backoff expiry timestamp)
 231          */
 232         private function syncBackoffs( array $backoffs ) {
 233                 $section = new ProfileSection( __METHOD__ );
 234
 235                 $file = wfTempDir() . '/mw-runJobs-backoffs.json';
 236                 $handle = fopen( $file, 'wb+' );
 237                 flock( $handle, LOCK_EX );
 238                 $content = stream_get_contents( $handle );
 239                 $cBackoffs = json_decode( $content, true ) ? : array();
 240                 foreach ( $backoffs as $type => $timestamp ) {
 241                         $cBackoffs[$type] = isset( $cBackoffs[$type] ) ? $cBackoffs[$type] : 0;
 242                         $cBackoffs[$type] = max( $cBackoffs[$type], $backoffs[$type] );
 243                 }
 244                 ftruncate( $handle, 0 );
 245                 fwrite( $handle, json_encode( $backoffs ) );
 246                 flock( $handle, LOCK_UN );
 247                 fclose( $handle );
 248         }
 249
 250         /**
 251          * Make sure that this script is not too close to the memory usage limit.
 252          * It is better to die in between jobs than OOM right in the middle of one.
 253          * @throws MWException
 254          */
 255         private function assertMemoryOK() {
 256                 static $maxBytes = null;
 257                 if ( $maxBytes === null ) {
 258                         $m = array();
 259                         if ( preg_match( '!^(\d+)(k|m|g|)$!i', ini_get( 'memory_limit' ), $m ) ) {
 260                                 list( , $num, $unit ) = $m;
 261                                 $conv = array( 'g' => 1073741824, 'm' => 1048576, 'k' => 1024, '' => 1 );
 262                                 $maxBytes = $num * $conv[strtolower( $unit )];
 263                         } else {
 264                                 $maxBytes = 0;
 265                         }
 266                 }
 267                 $usedBytes = memory_get_usage();
 268                 if ( $maxBytes && $usedBytes >= 0.95 * $maxBytes ) {
 269                         throw new MWException( "Detected excessive memory usage ($usedBytes/$maxBytes)." );
 270                 }
 271         }
 272
 273         /**
 274          * Log the job message
 275          * @param string $msg The message to log
 276          */
 277         private function runJobsLog( $msg ) {
 278                 $this->output( wfTimestamp( TS_DB ) . " $msg\n" );
 279                 wfDebugLog( 'runJobs', $msg );
 280         }
 281 }
 282
 283 $maintClass = "RunJobs";
 284 require_once RUN_MAINTENANCE_IF_MAIN;