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