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