includes/deferred/CdnCacheUpdate.php

   1 <?php
   2 /**
   3  * CDN cache purging.
   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  */
  22
  23 use Wikimedia\Assert\Assert;
  24 use MediaWiki\MediaWikiServices;
  25
  26 /**
  27  * Handles purging appropriate CDN URLs given a title (or titles)
  28  * @ingroup Cache
  29  */
  30 class CdnCacheUpdate implements DeferrableUpdate, MergeableUpdate {
  31         /** @var string[] Collection of URLs to purge */
  32         protected $urls = [];
  33
  34         /**
  35          * @param string[] $urlArr Collection of URLs to purge
  36          */
  37         public function __construct( array $urlArr ) {
  38                 $this->urls = $urlArr;
  39         }
  40
  41         public function merge( MergeableUpdate $update ) {
  42                 /** @var CdnCacheUpdate $update */
  43                 Assert::parameterType( __CLASS__, $update, '$update' );
  44
  45                 $this->urls = array_merge( $this->urls, $update->urls );
  46         }
  47
  48         /**
  49          * Create an update object from an array of Title objects, or a TitleArray object
  50          *
  51          * @param Traversable|Title[] $titles
  52          * @param string[] $urlArr
  53          * @return CdnCacheUpdate
  54          */
  55         public static function newFromTitles( $titles, $urlArr = [] ) {
  56                 ( new LinkBatch( $titles ) )->execute();
  57                 /** @var Title $title */
  58                 foreach ( $titles as $title ) {
  59                         $urlArr = array_merge( $urlArr, $title->getCdnUrls() );
  60                 }
  61
  62                 return new CdnCacheUpdate( $urlArr );
  63         }
  64
  65         /**
  66          * @param Title $title
  67          * @return CdnCacheUpdate
  68          * @deprecated since 1.27
  69          */
  70         public static function newSimplePurge( Title $title ) {
  71                 return new CdnCacheUpdate( $title->getCdnUrls() );
  72         }
  73
  74         /**
  75          * Purges the list of URLs passed to the constructor.
  76          */
  77         public function doUpdate() {
  78                 global $wgCdnReboundPurgeDelay;
  79
  80                 self::purge( $this->urls );
  81
  82                 if ( $wgCdnReboundPurgeDelay > 0 ) {
  83                         JobQueueGroup::singleton()->lazyPush( new CdnPurgeJob(
  84                                 Title::makeTitle( NS_SPECIAL, 'Badtitle/' . __CLASS__ ),
  85                                 [
  86                                         'urls' => $this->urls,
  87                                         'jobReleaseTimestamp' => time() + $wgCdnReboundPurgeDelay
  88                                 ]
  89                         ) );
  90                 }
  91         }
  92
  93         /**
  94          * Purges a list of CDN nodes defined in $wgSquidServers.
  95          * $urlArr should contain the full URLs to purge as values
  96          * (example: $urlArr[] = 'http://my.host/something')
  97          *
  98          * @param string[] $urlArr List of full URLs to purge
  99          */
 100         public static function purge( array $urlArr ) {
 101                 global $wgSquidServers, $wgHTCPRouting;
 102
 103                 if ( !$urlArr ) {
 104                         return;
 105                 }
 106
 107                 // Remove duplicate URLs from list
 108                 $urlArr = array_unique( $urlArr );
 109
 110                 wfDebugLog( 'squid', __METHOD__ . ': ' . implode( ' ', $urlArr ) );
 111
 112                 // Reliably broadcast the purge to all edge nodes
 113                 $relayer = MediaWikiServices::getInstance()->getEventRelayerGroup()
 114                                         ->getRelayer( 'cdn-url-purges' );
 115                 $ts = microtime( true );
 116                 $relayer->notifyMulti(
 117                         'cdn-url-purges',
 118                         array_map(
 119                                 function ( $url ) use ( $ts ) {
 120                                         return [
 121                                                 'url' => $url,
 122                                                 'timestamp' => $ts,
 123                                         ];
 124                                 },
 125                                 $urlArr
 126                         )
 127                 );
 128
 129                 // Send lossy UDP broadcasting if enabled
 130                 if ( $wgHTCPRouting ) {
 131                         self::HTCPPurge( $urlArr );
 132                 }
 133
 134                 // Do direct server purges if enabled (this does not scale very well)
 135                 if ( $wgSquidServers ) {
 136                         // Maximum number of parallel connections per squid
 137                         $maxSocketsPerSquid = 8;
 138                         // Number of requests to send per socket
 139                         // 400 seems to be a good tradeoff, opening a socket takes a while
 140                         $urlsPerSocket = 400;
 141                         $socketsPerSquid = ceil( count( $urlArr ) / $urlsPerSocket );
 142                         if ( $socketsPerSquid > $maxSocketsPerSquid ) {
 143                                 $socketsPerSquid = $maxSocketsPerSquid;
 144                         }
 145
 146                         $pool = new SquidPurgeClientPool;
 147                         $chunks = array_chunk( $urlArr, ceil( count( $urlArr ) / $socketsPerSquid ) );
 148                         foreach ( $wgSquidServers as $server ) {
 149                                 foreach ( $chunks as $chunk ) {
 150                                         $client = new SquidPurgeClient( $server );
 151                                         foreach ( $chunk as $url ) {
 152                                                 $client->queuePurge( $url );
 153                                         }
 154                                         $pool->addClient( $client );
 155                                 }
 156                         }
 157
 158                         $pool->run();
 159                 }
 160         }
 161
 162         /**
 163          * Send Hyper Text Caching Protocol (HTCP) CLR requests.
 164          *
 165          * @throws MWException
 166          * @param string[] $urlArr Collection of URLs to purge
 167          */
 168         private static function HTCPPurge( array $urlArr ) {
 169                 global $wgHTCPRouting, $wgHTCPMulticastTTL;
 170
 171                 // HTCP CLR operation
 172                 $htcpOpCLR = 4;
 173
 174                 // @todo FIXME: PHP doesn't support these socket constants (include/linux/in.h)
 175                 if ( !defined( "IPPROTO_IP" ) ) {
 176                         define( "IPPROTO_IP", 0 );
 177                         define( "IP_MULTICAST_LOOP", 34 );
 178                         define( "IP_MULTICAST_TTL", 33 );
 179                 }
 180
 181                 // pfsockopen doesn't work because we need set_sock_opt
 182                 $conn = socket_create( AF_INET, SOCK_DGRAM, SOL_UDP );
 183                 if ( !$conn ) {
 184                         $errstr = socket_strerror( socket_last_error() );
 185                         wfDebugLog( 'squid', __METHOD__ .
 186                                 ": Error opening UDP socket: $errstr" );
 187
 188                         return;
 189                 }
 190
 191                 // Set socket options
 192                 socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_LOOP, 0 );
 193                 if ( $wgHTCPMulticastTTL != 1 ) {
 194                         // Set multicast time to live (hop count) option on socket
 195                         socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_TTL,
 196                                 $wgHTCPMulticastTTL );
 197                 }
 198
 199                 // Get sequential trx IDs for packet loss counting
 200                 $ids = UIDGenerator::newSequentialPerNodeIDs(
 201                         'squidhtcppurge', 32, count( $urlArr ), UIDGenerator::QUICK_VOLATILE
 202                 );
 203
 204                 foreach ( $urlArr as $url ) {
 205                         if ( !is_string( $url ) ) {
 206                                 throw new MWException( 'Bad purge URL' );
 207                         }
 208                         $url = self::expand( $url );
 209                         $conf = self::getRuleForURL( $url, $wgHTCPRouting );
 210                         if ( !$conf ) {
 211                                 wfDebugLog( 'squid', __METHOD__ .
 212                                         "No HTCP rule configured for URL {$url} , skipping" );
 213                                 continue;
 214                         }
 215
 216                         if ( isset( $conf['host'] ) && isset( $conf['port'] ) ) {
 217                                 // Normalize single entries
 218                                 $conf = [ $conf ];
 219                         }
 220                         foreach ( $conf as $subconf ) {
 221                                 if ( !isset( $subconf['host'] ) || !isset( $subconf['port'] ) ) {
 222                                         throw new MWException( "Invalid HTCP rule for URL $url\n" );
 223                                 }
 224                         }
 225
 226                         // Construct a minimal HTCP request diagram
 227                         // as per RFC 2756
 228                         // Opcode 'CLR', no response desired, no auth
 229                         $htcpTransID = current( $ids );
 230                         next( $ids );
 231
 232                         $htcpSpecifier = pack( 'na4na*na8n',
 233                                 4, 'HEAD', strlen( $url ), $url,
 234                                 8, 'HTTP/1.0', 0 );
 235
 236                         $htcpDataLen = 8 + 2 + strlen( $htcpSpecifier );
 237                         $htcpLen = 4 + $htcpDataLen + 2;
 238
 239                         // Note! Squid gets the bit order of the first
 240                         // word wrong, wrt the RFC. Apparently no other
 241                         // implementation exists, so adapt to Squid
 242                         $htcpPacket = pack( 'nxxnCxNxxa*n',
 243                                 $htcpLen, $htcpDataLen, $htcpOpCLR,
 244                                 $htcpTransID, $htcpSpecifier, 2 );
 245
 246                         wfDebugLog( 'squid', __METHOD__ .
 247                                 "Purging URL $url via HTCP" );
 248                         foreach ( $conf as $subconf ) {
 249                                 socket_sendto( $conn, $htcpPacket, $htcpLen, 0,
 250                                         $subconf['host'], $subconf['port'] );
 251                         }
 252                 }
 253         }
 254
 255         /**
 256          * Expand local URLs to fully-qualified URLs using the internal protocol
 257          * and host defined in $wgInternalServer. Input that's already fully-
 258          * qualified will be passed through unchanged.
 259          *
 260          * This is used to generate purge URLs that may be either local to the
 261          * main wiki or include a non-native host, such as images hosted on a
 262          * second internal server.
 263          *
 264          * Client functions should not need to call this.
 265          *
 266          * @param string $url
 267          * @return string
 268          */
 269         public static function expand( $url ) {
 270                 return wfExpandUrl( $url, PROTO_INTERNAL );
 271         }
 272
 273         /**
 274          * Find the HTCP routing rule to use for a given URL.
 275          * @param string $url URL to match
 276          * @param array $rules Array of rules, see $wgHTCPRouting for format and behavior
 277          * @return mixed Element of $rules that matched, or false if nothing matched
 278          */
 279         private static function getRuleForURL( $url, $rules ) {
 280                 foreach ( $rules as $regex => $routing ) {
 281                         if ( $regex === '' || preg_match( $regex, $url ) ) {
 282                                 return $routing;
 283                         }
 284                 }
 285
 286                 return false;
 287         }
 288 }
 289
 290 /**
 291  * @deprecated since 1.27
 292  */
 293 class SquidUpdate extends CdnCacheUpdate {
 294         // Keep class name for b/c
 295 }