includes/cache/SquidUpdate.php

   1 <?php
   2 /**
   3  * Squid 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 /**
  25  * Handles purging appropriate Squid URLs given a title (or titles)
  26  * @ingroup Cache
  27  */
  28 class SquidUpdate {
  29         var $urlArr, $mMaxTitles;
  30
  31         function __construct( $urlArr = Array(), $maxTitles = false ) {
  32                 global $wgMaxSquidPurgeTitles;
  33                 if ( $maxTitles === false ) {
  34                         $this->mMaxTitles = $wgMaxSquidPurgeTitles;
  35                 } else {
  36                         $this->mMaxTitles = $maxTitles;
  37                 }
  38                 $urlArr = array_unique( $urlArr ); // Remove duplicates
  39                 if ( count( $urlArr ) > $this->mMaxTitles ) {
  40                         $urlArr = array_slice( $urlArr, 0, $this->mMaxTitles );
  41                 }
  42                 $this->urlArr = $urlArr;
  43         }
  44
  45         /**
  46          * @param $title Title
  47          *
  48          * @return SquidUpdate
  49          */
  50         static function newFromLinksTo( &$title ) {
  51                 global $wgMaxSquidPurgeTitles;
  52                 wfProfileIn( __METHOD__ );
  53
  54                 # Get a list of URLs linking to this page
  55                 $dbr = wfGetDB( DB_SLAVE );
  56                 $res = $dbr->select( array( 'links', 'page' ),
  57                         array( 'page_namespace', 'page_title' ),
  58                         array(
  59                                 'pl_namespace' => $title->getNamespace(),
  60                                 'pl_title'     => $title->getDBkey(),
  61                                 'pl_from=page_id' ),
  62                         __METHOD__ );
  63                 $blurlArr = $title->getSquidURLs();
  64                 if ( $dbr->numRows( $res ) <= $wgMaxSquidPurgeTitles ) {
  65                         foreach ( $res as $BL ) {
  66                                 $tobj = Title::makeTitle( $BL->page_namespace, $BL->page_title ) ;
  67                                 $blurlArr[] = $tobj->getInternalURL();
  68                         }
  69                 }
  70
  71                 wfProfileOut( __METHOD__ );
  72                 return new SquidUpdate( $blurlArr );
  73         }
  74
  75         /**
  76          * Create a SquidUpdate from an array of Title objects, or a TitleArray object
  77          *
  78          * @param $titles array
  79          * @param $urlArr array
  80          *
  81          * @return SquidUpdate
  82          */
  83         static function newFromTitles( $titles, $urlArr = array() ) {
  84                 global $wgMaxSquidPurgeTitles;
  85                 $i = 0;
  86                 foreach ( $titles as $title ) {
  87                         $urlArr[] = $title->getInternalURL();
  88                         if ( $i++ > $wgMaxSquidPurgeTitles ) {
  89                                 break;
  90                         }
  91                 }
  92                 return new SquidUpdate( $urlArr );
  93         }
  94
  95         /**
  96          * @param $title Title
  97          *
  98          * @return SquidUpdate
  99          */
 100         static function newSimplePurge( &$title ) {
 101                 $urlArr = $title->getSquidURLs();
 102                 return new SquidUpdate( $urlArr );
 103         }
 104
 105         /**
 106          * Purges the list of URLs passed to the constructor
 107          */
 108         function doUpdate() {
 109                 SquidUpdate::purge( $this->urlArr );
 110         }
 111
 112         /**
 113          * Purges a list of Squids defined in $wgSquidServers.
 114          * $urlArr should contain the full URLs to purge as values
 115          * (example: $urlArr[] = 'http://my.host/something')
 116          * XXX report broken Squids per mail or log
 117          *
 118          * @param $urlArr array
 119          * @return void
 120          */
 121         static function purge( $urlArr ) {
 122                 global $wgSquidServers, $wgHTCPMulticastRouting;
 123
 124                 /*if ( (@$wgSquidServers[0]) == 'echo' ) {
 125                         echo implode("<br />\n", $urlArr) . "<br />\n";
 126                         return;
 127                 }*/
 128
 129                 if( !$urlArr ) {
 130                         return;
 131                 }
 132
 133                 if ( $wgHTCPMulticastRouting ) {
 134                         SquidUpdate::HTCPPurge( $urlArr );
 135                 }
 136
 137                 wfProfileIn( __METHOD__ );
 138
 139                 $urlArr = array_unique( $urlArr ); // Remove duplicates
 140                 $maxSocketsPerSquid = 8; //  socket cap per Squid
 141                 $urlsPerSocket = 400; // 400 seems to be a good tradeoff, opening a socket takes a while
 142                 $socketsPerSquid = ceil( count( $urlArr ) / $urlsPerSocket );
 143                 if ( $socketsPerSquid > $maxSocketsPerSquid ) {
 144                         $socketsPerSquid = $maxSocketsPerSquid;
 145                 }
 146
 147                 $pool = new SquidPurgeClientPool;
 148                 $chunks = array_chunk( $urlArr, ceil( count( $urlArr ) / $socketsPerSquid ) );
 149                 foreach ( $wgSquidServers as $server ) {
 150                         foreach ( $chunks as $chunk ) {
 151                                 $client = new SquidPurgeClient( $server );
 152                                 foreach ( $chunk as $url ) {
 153                                         $client->queuePurge( $url );
 154                                 }
 155                                 $pool->addClient( $client );
 156                         }
 157                 }
 158                 $pool->run();
 159
 160                 wfProfileOut( __METHOD__ );
 161         }
 162
 163         /**
 164          * @throws MWException
 165          * @param $urlArr array
 166          */
 167         static function HTCPPurge( $urlArr ) {
 168                 global $wgHTCPMulticastRouting, $wgHTCPMulticastTTL;
 169                 wfProfileIn( __METHOD__ );
 170
 171                 $htcpOpCLR = 4; // HTCP CLR
 172
 173                 // @todo FIXME: PHP doesn't support these socket constants (include/linux/in.h)
 174                 if( !defined( "IPPROTO_IP" ) ) {
 175                         define( "IPPROTO_IP", 0 );
 176                         define( "IP_MULTICAST_LOOP", 34 );
 177                         define( "IP_MULTICAST_TTL", 33 );
 178                 }
 179
 180                 // pfsockopen doesn't work because we need set_sock_opt
 181                 $conn = socket_create( AF_INET, SOCK_DGRAM, SOL_UDP );
 182                 if ( $conn ) {
 183                         // Set socket options
 184                         socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_LOOP, 0 );
 185                         if ( $wgHTCPMulticastTTL != 1 )
 186                                 socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_TTL,
 187                                         $wgHTCPMulticastTTL );
 188
 189                         $urlArr = array_unique( $urlArr ); // Remove duplicates
 190                         foreach ( $urlArr as $url ) {
 191                                 if( !is_string( $url ) ) {
 192                                         throw new MWException( 'Bad purge URL' );
 193                                 }
 194                                 $url = SquidUpdate::expand( $url );
 195                                 $conf = self::getRuleForURL( $url, $wgHTCPMulticastRouting );
 196                                 if ( !$conf ) {
 197                                         wfDebug( "No HTCP rule configured for URL $url , skipping\n" );
 198                                         continue;
 199                                 }
 200                                 if ( !isset( $conf['host'] ) || !isset( $conf['port'] ) ) {
 201                                         throw new MWException( "Invalid HTCP rule for URL $url\n" );
 202                                 }
 203
 204                                 // Construct a minimal HTCP request diagram
 205                                 // as per RFC 2756
 206                                 // Opcode 'CLR', no response desired, no auth
 207                                 $htcpTransID = rand();
 208
 209                                 $htcpSpecifier = pack( 'na4na*na8n',
 210                                         4, 'HEAD', strlen( $url ), $url,
 211                                         8, 'HTTP/1.0', 0 );
 212
 213                                 $htcpDataLen = 8 + 2 + strlen( $htcpSpecifier );
 214                                 $htcpLen = 4 + $htcpDataLen + 2;
 215
 216                                 // Note! Squid gets the bit order of the first
 217                                 // word wrong, wrt the RFC. Apparently no other
 218                                 // implementation exists, so adapt to Squid
 219                                 $htcpPacket = pack( 'nxxnCxNxxa*n',
 220                                         $htcpLen, $htcpDataLen, $htcpOpCLR,
 221                                         $htcpTransID, $htcpSpecifier, 2);
 222
 223                                 // Send out
 224                                 wfDebug( "Purging URL $url via HTCP\n" );
 225                                 socket_sendto( $conn, $htcpPacket, $htcpLen, 0,
 226                                         $conf['host'], $conf['port'] );
 227                         }
 228                 } else {
 229                         $errstr = socket_strerror( socket_last_error() );
 230                         wfDebug( __METHOD__ . "(): Error opening UDP socket: $errstr\n" );
 231                 }
 232                 wfProfileOut( __METHOD__ );
 233         }
 234
 235         /**
 236          * Expand local URLs to fully-qualified URLs using the internal protocol
 237          * and host defined in $wgInternalServer. Input that's already fully-
 238          * qualified will be passed through unchanged.
 239          *
 240          * This is used to generate purge URLs that may be either local to the
 241          * main wiki or include a non-native host, such as images hosted on a
 242          * second internal server.
 243          *
 244          * Client functions should not need to call this.
 245          *
 246          * @param $url string
 247          *
 248          * @return string
 249          */
 250         static function expand( $url ) {
 251                 return wfExpandUrl( $url, PROTO_INTERNAL );
 252         }
 253
 254         /**
 255          * Find the HTCP routing rule to use for a given URL.
 256          * @param $url string URL to match
 257          * @param $rules array Array of rules, see $wgHTCPMulticastRouting for format and behavior
 258          * @return mixed Element of $rules that matched, or false if nothing matched
 259          */
 260         static function getRuleForURL( $url, $rules ) {
 261                 foreach ( $rules as $regex => $routing ) {
 262                         if ( $regex === '' || preg_match( $regex, $url ) ) {
 263                                 return $routing;
 264                         }
 265                 }
 266                 return false;
 267         }
 268
 269 }