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         /** @var string[] Collection of URLs to purge */
  30         protected $urls = array();
  31
  32         /**
  33          * @param array $urlArr Collection of URLs to purge
  34          */
  35         public function __construct( array $urlArr ) {
  36                 // Remove duplicate URLs from list
  37                 $this->urls = array_unique( $urlArr );
  38         }
  39
  40         /**
  41          * Create a SquidUpdate from an array of Title objects, or a TitleArray object
  42          *
  43          * @param Traversable|array $titles
  44          * @param array $urlArr
  45          * @return SquidUpdate
  46          */
  47         public static function newFromTitles( $titles, $urlArr = array() ) {
  48                 /** @var Title $title */
  49                 foreach ( $titles as $title ) {
  50                         $urlArr[] = $title->getInternalURL();
  51                 }
  52
  53                 return new SquidUpdate( $urlArr );
  54         }
  55
  56         /**
  57          * @param Title $title
  58          * @return SquidUpdate
  59          * @deprecated 1.27
  60          */
  61         public static function newSimplePurge( Title $title ) {
  62                 $urlArr = $title->getSquidURLs();
  63
  64                 return new SquidUpdate( $urlArr );
  65         }
  66
  67         /**
  68          * Purges the list of URLs passed to the constructor.
  69          */
  70         public function doUpdate() {
  71                 self::purge( $this->urls );
  72         }
  73
  74         /**
  75          * Purges a list of Squids defined in $wgSquidServers.
  76          * $urlArr should contain the full URLs to purge as values
  77          * (example: $urlArr[] = 'http://my.host/something')
  78          * XXX report broken Squids per mail or log
  79          *
  80          * @param array $urlArr List of full URLs to purge
  81          */
  82         public static function purge( array $urlArr ) {
  83                 global $wgSquidServers, $wgHTCPRouting;
  84
  85                 if ( !$urlArr ) {
  86                         return;
  87                 }
  88
  89                 wfDebugLog( 'squid', __METHOD__ . ': ' . implode( ' ', $urlArr ) );
  90
  91                 if ( $wgHTCPRouting ) {
  92                         self::HTCPPurge( $urlArr );
  93                 }
  94
  95                 if ( $wgSquidServers ) {
  96                         // Remove duplicate URLs
  97                         $urlArr = array_unique( $urlArr );
  98                         // Maximum number of parallel connections per squid
  99                         $maxSocketsPerSquid = 8;
 100                         // Number of requests to send per socket
 101                         // 400 seems to be a good tradeoff, opening a socket takes a while
 102                         $urlsPerSocket = 400;
 103                         $socketsPerSquid = ceil( count( $urlArr ) / $urlsPerSocket );
 104                         if ( $socketsPerSquid > $maxSocketsPerSquid ) {
 105                                 $socketsPerSquid = $maxSocketsPerSquid;
 106                         }
 107
 108                         $pool = new SquidPurgeClientPool;
 109                         $chunks = array_chunk( $urlArr, ceil( count( $urlArr ) / $socketsPerSquid ) );
 110                         foreach ( $wgSquidServers as $server ) {
 111                                 foreach ( $chunks as $chunk ) {
 112                                         $client = new SquidPurgeClient( $server );
 113                                         foreach ( $chunk as $url ) {
 114                                                 $client->queuePurge( $url );
 115                                         }
 116                                         $pool->addClient( $client );
 117                                 }
 118                         }
 119
 120                         $pool->run();
 121                 }
 122         }
 123
 124         /**
 125          * Send Hyper Text Caching Protocol (HTCP) CLR requests.
 126          *
 127          * @throws MWException
 128          * @param array $urlArr Collection of URLs to purge
 129          */
 130         protected static function HTCPPurge( $urlArr ) {
 131                 global $wgHTCPRouting, $wgHTCPMulticastTTL;
 132
 133                 // HTCP CLR operation
 134                 $htcpOpCLR = 4;
 135
 136                 // @todo FIXME: PHP doesn't support these socket constants (include/linux/in.h)
 137                 if ( !defined( "IPPROTO_IP" ) ) {
 138                         define( "IPPROTO_IP", 0 );
 139                         define( "IP_MULTICAST_LOOP", 34 );
 140                         define( "IP_MULTICAST_TTL", 33 );
 141                 }
 142
 143                 // pfsockopen doesn't work because we need set_sock_opt
 144                 $conn = socket_create( AF_INET, SOCK_DGRAM, SOL_UDP );
 145                 if ( !$conn ) {
 146                         $errstr = socket_strerror( socket_last_error() );
 147                         wfDebugLog( 'squid', __METHOD__ .
 148                                 ": Error opening UDP socket: $errstr" );
 149
 150                         return;
 151                 }
 152
 153                 // Set socket options
 154                 socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_LOOP, 0 );
 155                 if ( $wgHTCPMulticastTTL != 1 ) {
 156                         // Set multicast time to live (hop count) option on socket
 157                         socket_set_option( $conn, IPPROTO_IP, IP_MULTICAST_TTL,
 158                                 $wgHTCPMulticastTTL );
 159                 }
 160
 161                 // Remove duplicate URLs from collection
 162                 $urlArr = array_unique( $urlArr );
 163                 // Get sequential trx IDs for packet loss counting
 164                 $ids = UIDGenerator::newSequentialPerNodeIDs(
 165                         'squidhtcppurge', 32, count( $urlArr ), UIDGenerator::QUICK_VOLATILE
 166                 );
 167
 168                 foreach ( $urlArr as $url ) {
 169                         if ( !is_string( $url ) ) {
 170                                 throw new MWException( 'Bad purge URL' );
 171                         }
 172                         $url = self::expand( $url );
 173                         $conf = self::getRuleForURL( $url, $wgHTCPRouting );
 174                         if ( !$conf ) {
 175                                 wfDebugLog( 'squid', __METHOD__ .
 176                                         "No HTCP rule configured for URL {$url} , skipping" );
 177                                 continue;
 178                         }
 179
 180                         if ( isset( $conf['host'] ) && isset( $conf['port'] ) ) {
 181                                 // Normalize single entries
 182                                 $conf = array( $conf );
 183                         }
 184                         foreach ( $conf as $subconf ) {
 185                                 if ( !isset( $subconf['host'] ) || !isset( $subconf['port'] ) ) {
 186                                         throw new MWException( "Invalid HTCP rule for URL $url\n" );
 187                                 }
 188                         }
 189
 190                         // Construct a minimal HTCP request diagram
 191                         // as per RFC 2756
 192                         // Opcode 'CLR', no response desired, no auth
 193                         $htcpTransID = current( $ids );
 194                         next( $ids );
 195
 196                         $htcpSpecifier = pack( 'na4na*na8n',
 197                                 4, 'HEAD', strlen( $url ), $url,
 198                                 8, 'HTTP/1.0', 0 );
 199
 200                         $htcpDataLen = 8 + 2 + strlen( $htcpSpecifier );
 201                         $htcpLen = 4 + $htcpDataLen + 2;
 202
 203                         // Note! Squid gets the bit order of the first
 204                         // word wrong, wrt the RFC. Apparently no other
 205                         // implementation exists, so adapt to Squid
 206                         $htcpPacket = pack( 'nxxnCxNxxa*n',
 207                                 $htcpLen, $htcpDataLen, $htcpOpCLR,
 208                                 $htcpTransID, $htcpSpecifier, 2 );
 209
 210                         wfDebugLog( 'squid', __METHOD__ .
 211                                 "Purging URL $url via HTCP" );
 212                         foreach ( $conf as $subconf ) {
 213                                 socket_sendto( $conn, $htcpPacket, $htcpLen, 0,
 214                                         $subconf['host'], $subconf['port'] );
 215                         }
 216                 }
 217         }
 218
 219         /**
 220          * Expand local URLs to fully-qualified URLs using the internal protocol
 221          * and host defined in $wgInternalServer. Input that's already fully-
 222          * qualified will be passed through unchanged.
 223          *
 224          * This is used to generate purge URLs that may be either local to the
 225          * main wiki or include a non-native host, such as images hosted on a
 226          * second internal server.
 227          *
 228          * Client functions should not need to call this.
 229          *
 230          * @param string $url
 231          * @return string
 232          */
 233         public static function expand( $url ) {
 234                 return wfExpandUrl( $url, PROTO_INTERNAL );
 235         }
 236
 237         /**
 238          * Find the HTCP routing rule to use for a given URL.
 239          * @param string $url URL to match
 240          * @param array $rules Array of rules, see $wgHTCPRouting for format and behavior
 241          * @return mixed Element of $rules that matched, or false if nothing matched
 242          */
 243         private static function getRuleForURL( $url, $rules ) {
 244                 foreach ( $rules as $regex => $routing ) {
 245                         if ( $regex === '' || preg_match( $regex, $url ) ) {
 246                                 return $routing;
 247                         }
 248                 }
 249
 250                 return false;
 251         }
 252 }