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