includes/job/JobQueue.php

   1 <?php
   2 /**
   3  * Job queue base code.
   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  * @defgroup JobQueue JobQueue
  22  * @author Aaron Schulz
  23  */
  24
  25 /**
  26  * Class to handle enqueueing and running of background jobs
  27  *
  28  * @ingroup JobQueue
  29  * @since 1.21
  30  */
  31 abstract class JobQueue {
  32         protected $wiki; // string; wiki ID
  33         protected $type; // string; job type
  34         protected $order; // string; job priority for pop()
  35         protected $claimTTL; // integer; seconds
  36
  37         const QoS_Atomic = 1; // integer; "all-or-nothing" job insertions
  38
  39         /**
  40          * @param $params array
  41          */
  42         protected function __construct( array $params ) {
  43                 $this->wiki     = $params['wiki'];
  44                 $this->type     = $params['type'];
  45                 $this->order    = isset( $params['order'] ) ? $params['order'] : 'random';
  46                 $this->claimTTL = isset( $params['claimTTL'] ) ? $params['claimTTL'] : 0;
  47         }
  48
  49         /**
  50          * Get a job queue object of the specified type.
  51          * $params includes:
  52          *   class    : What job class to use (determines job type)
  53          *   wiki     : wiki ID of the wiki the jobs are for (defaults to current wiki)
  54          *   type     : The name of the job types this queue handles
  55          *   order    : Order that pop() selects jobs, one of "fifo", "timestamp" or "random".
  56          *              If "fifo" is used, the queue will effectively be FIFO. Note that
  57          *              job completion will not appear to be exactly FIFO if there are multiple
  58          *              job runners since jobs can take different times to finish once popped.
  59          *              If "timestamp" is used, the queue will at least be loosely ordered
  60          *              by timestamp, allowing for some jobs to be popped off out of order.
  61          *              If "random" is used, pop() will pick jobs in random order. This might be
  62          *              useful for improving concurrency depending on the queue storage medium.
  63          *   claimTTL : If supported, the queue will recycle jobs that have been popped
  64          *              but not acknowledged as completed after this many seconds.
  65          *
  66          * @param $params array
  67          * @return JobQueue
  68          * @throws MWException
  69          */
  70         final public static function factory( array $params ) {
  71                 $class = $params['class'];
  72                 if ( !MWInit::classExists( $class ) ) {
  73                         throw new MWException( "Invalid job queue class '$class'." );
  74                 }
  75                 $obj = new $class( $params );
  76                 if ( !( $obj instanceof self ) ) {
  77                         throw new MWException( "Class '$class' is not a " . __CLASS__ . " class." );
  78                 }
  79                 return $obj;
  80         }
  81
  82         /**
  83          * @return string Wiki ID
  84          */
  85         final public function getWiki() {
  86                 return $this->wiki;
  87         }
  88
  89         /**
  90          * @return string Job type that this queue handles
  91          */
  92         final public function getType() {
  93                 return $this->type;
  94         }
  95
  96         /**
  97          * Quickly check if the queue is empty.
  98          * Queue classes should use caching if they are any slower without memcached.
  99          *
 100          * @return bool
 101          */
 102         final public function isEmpty() {
 103                 wfProfileIn( __METHOD__ );
 104                 $res = $this->doIsEmpty();
 105                 wfProfileOut( __METHOD__ );
 106                 return $res;
 107         }
 108
 109         /**
 110          * @see JobQueue::isEmpty()
 111          * @return bool
 112          */
 113         abstract protected function doIsEmpty();
 114
 115         /**
 116          * Push a batch of jobs into the queue
 117          *
 118          * @param $jobs array List of Jobs
 119          * @param $flags integer Bitfield (supports JobQueue::QoS_Atomic)
 120          * @throws MWException
 121          * @return bool
 122          */
 123         final public function batchPush( array $jobs, $flags = 0 ) {
 124                 foreach ( $jobs as $job ) {
 125                         if ( $job->getType() !== $this->type ) {
 126                                 throw new MWException( "Got '{$job->getType()}' job; expected '{$this->type}'." );
 127                         }
 128                 }
 129                 wfProfileIn( __METHOD__ );
 130                 $ok = $this->doBatchPush( $jobs, $flags );
 131                 wfProfileOut( __METHOD__ );
 132                 return $ok;
 133         }
 134
 135         /**
 136          * @see JobQueue::batchPush()
 137          * @return bool
 138          */
 139         abstract protected function doBatchPush( array $jobs, $flags );
 140
 141         /**
 142          * Pop a job off of the queue
 143          *
 144          * @return Job|bool Returns false on failure
 145          */
 146         final public function pop() {
 147                 wfProfileIn( __METHOD__ );
 148                 $job = $this->doPop();
 149                 wfProfileOut( __METHOD__ );
 150                 return $job;
 151         }
 152
 153         /**
 154          * @see JobQueue::pop()
 155          * @return Job
 156          */
 157         abstract protected function doPop();
 158
 159         /**
 160          * Acknowledge that a job was completed
 161          *
 162          * @param $job Job
 163          * @throws MWException
 164          * @return bool
 165          */
 166         final public function ack( Job $job ) {
 167                 if ( $job->getType() !== $this->type ) {
 168                         throw new MWException( "Got '{$job->getType()}' job; expected '{$this->type}'." );
 169                 }
 170                 wfProfileIn( __METHOD__ );
 171                 $ok = $this->doAck( $job );
 172                 wfProfileOut( __METHOD__ );
 173                 return $ok;
 174         }
 175
 176         /**
 177          * @see JobQueue::ack()
 178          * @return bool
 179          */
 180         abstract protected function doAck( Job $job );
 181
 182         /**
 183          * Register the "root job" of a given job into the queue for de-duplication.
 184          * This should only be called right *after* all the new jobs have been inserted.
 185          * This is used to turn older, duplicate, job entries into no-ops. The root job
 186          * information will remain in the registry until it simply falls out of cache.
 187          *
 188          * This requires that $job has two special fields in the "params" array:
 189          *   - rootJobSignature : hash (e.g. SHA1) that identifies the task
 190          *   - rootJobTimestamp : TS_MW timestamp of this instance of the task
 191          *
 192          * A "root job" is a conceptual job that consist of potentially many smaller jobs
 193          * that are actually inserted into the queue. For example, "refreshLinks" jobs are
 194          * spawned when a template is edited. One can think of the task as "update links
 195          * of pages that use template X" and an instance of that task as a "root job".
 196          * However, what actually goes into the queue are potentially many refreshLinks2 jobs.
 197          * Since these jobs include things like page ID ranges and DB master positions, and morph
 198          * into smaller refreshLinks2 jobs recursively, simple duplicate detection (like job_sha1)
 199          * for individual jobs being identical is not useful.
 200          *
 201          * In the case of "refreshLinks", if these jobs are still in the queue when the template
 202          * is edited again, we want all of these old refreshLinks jobs for that template to become
 203          * no-ops. This can greatly reduce server load, since refreshLinks jobs involves parsing.
 204          * Essentially, the new batch of jobs belong to a new "root job" and the older ones to a
 205          * previous "root job" for the same task of "update links of pages that use template X".
 206          *
 207          * @param $job Job
 208          * @throws MWException
 209          * @return bool
 210          */
 211         final public function deduplicateRootJob( Job $job ) {
 212                 if ( $job->getType() !== $this->type ) {
 213                         throw new MWException( "Got '{$job->getType()}' job; expected '{$this->type}'." );
 214                 }
 215                 wfProfileIn( __METHOD__ );
 216                 $ok = $this->doDeduplicateRootJob( $job );
 217                 wfProfileOut( __METHOD__ );
 218                 return $ok;
 219         }
 220
 221         /**
 222          * @see JobQueue::deduplicateRootJob()
 223          * @param $job Job
 224          * @return bool
 225          */
 226         protected function doDeduplicateRootJob( Job $job ) {
 227                 return true;
 228         }
 229
 230         /**
 231          * Wait for any slaves or backup servers to catch up
 232          *
 233          * @return void
 234          */
 235         final public function waitForBackups() {
 236                 wfProfileIn( __METHOD__ );
 237                 $this->doWaitForBackups();
 238                 wfProfileOut( __METHOD__ );
 239         }
 240
 241         /**
 242          * @see JobQueue::waitForBackups()
 243          * @return void
 244          */
 245         protected function doWaitForBackups() {}
 246 }