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