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