maintenance/fixSlaveDesync.php

   1 <?php
   2 /**
   3  * Fix erroneous page_latest values due to slave desynchronisation.
   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 Maintenance
  22  */
  23
  24 require_once __DIR__ . '/Maintenance.php';
  25
  26 /**
  27  * Maintenance script that fixes erroneous page_latest values
  28  * due to slave desynchronisation.
  29  *
  30  * @ingroup Maintenance
  31  */
  32 class FixSlaveDesync extends Maintenance {
  33         /** @var array */
  34         private $slaveIndexes;
  35
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->mDescription = "";
  39         }
  40
  41         public function getDbType() {
  42                 return Maintenance::DB_ADMIN;
  43         }
  44
  45         public function execute() {
  46                 $this->slaveIndexes = array();
  47                 $serverCount = wfGetLB()->getServerCount();
  48                 for ( $i = 1; $i < $serverCount; $i++ ) {
  49                         if ( wfGetLB()->isNonZeroLoad( $i ) ) {
  50                                 $this->slaveIndexes[] = $i;
  51                         }
  52                 }
  53
  54                 if ( $this->hasArg() ) {
  55                         $this->desyncFixPage( $this->getArg() );
  56                 } else {
  57                         $corrupt = $this->findPageLatestCorruption();
  58                         foreach ( $corrupt as $id => $dummy ) {
  59                                 $this->desyncFixPage( $id );
  60                         }
  61                 }
  62         }
  63
  64         /**
  65          * Find all pages that have a corrupted page_latest
  66          * @return array
  67          */
  68         private function findPageLatestCorruption() {
  69                 $desync = array();
  70                 $n = 0;
  71                 $dbw = wfGetDB( DB_MASTER );
  72                 $masterIDs = array();
  73                 $res = $dbw->select(
  74                         'page',
  75                         array( 'page_id', 'page_latest' ),
  76                         array( 'page_id<6054123' ),
  77                         __METHOD__
  78                 );
  79                 $this->output( "Number of pages: " . $res->numRows() . "\n" );
  80                 foreach ( $res as $row ) {
  81                         $masterIDs[$row->page_id] = $row->page_latest;
  82                         if ( !( ++$n % 10000 ) ) {
  83                                 $this->output( "$n\r" );
  84                         }
  85                 }
  86                 $this->output( "\n" );
  87
  88                 foreach ( $this->slaveIndexes as $i ) {
  89                         $db = wfGetDB( $i );
  90                         $res = $db->select(
  91                                 'page',
  92                                 array( 'page_id', 'page_latest' ),
  93                                 array( 'page_id<6054123' ),
  94                                 __METHOD__
  95                         );
  96                         foreach ( $res as $row ) {
  97                                 if ( isset( $masterIDs[$row->page_id] ) && $masterIDs[$row->page_id] != $row->page_latest ) {
  98                                         $desync[$row->page_id] = true;
  99                                         $this->output( $row->page_id . "\t" );
 100                                 }
 101                         }
 102                 }
 103                 $this->output( "\n" );
 104
 105                 return $desync;
 106         }
 107
 108         /**
 109          * Fix a broken page entry
 110          * @param int $pageID The page_id to fix
 111          */
 112         private function desyncFixPage( $pageID ) {
 113                 # Check for a corrupted page_latest
 114                 $dbw = wfGetDB( DB_MASTER );
 115                 $dbw->begin( __METHOD__ );
 116                 $realLatest = $dbw->selectField( 'page', 'page_latest', array( 'page_id' => $pageID ),
 117                         __METHOD__, 'FOR UPDATE' );
 118                 # list( $masterFile, $masterPos ) = $dbw->getMasterPos();
 119                 $found = false;
 120                 foreach ( $this->slaveIndexes as $i ) {
 121                         $db = wfGetDB( $i );
 122                         /*
 123                         if ( !$db->masterPosWait( $masterFile, $masterPos, 10 ) ) {
 124                                 $this->output( "Slave is too lagged, aborting\n" );
 125                                 $dbw->commit( __METHOD__ );
 126                                 sleep(10);
 127                                 return;
 128                         }*/
 129                         $latest = $db->selectField( 'page', 'page_latest', array( 'page_id' => $pageID ), __METHOD__ );
 130                         $max = $db->selectField( 'revision', 'MAX(rev_id)', false, __METHOD__ );
 131                         if ( $latest != $realLatest && $realLatest < $max ) {
 132                                 $this->output( "page_latest corrupted in page $pageID, server $i\n" );
 133                                 $found = true;
 134                                 break;
 135                         }
 136                 }
 137                 if ( !$found ) {
 138                         $this->output( "page_id $pageID seems fine\n" );
 139                         $dbw->commit( __METHOD__ );
 140
 141                         return;
 142                 }
 143
 144                 # Find the missing revisions
 145                 $res = $dbw->select( 'revision', array( 'rev_id' ), array( 'rev_page' => $pageID ),
 146                         __METHOD__, 'FOR UPDATE' );
 147                 $masterIDs = array();
 148                 foreach ( $res as $row ) {
 149                         $masterIDs[] = $row->rev_id;
 150                 }
 151
 152                 $res = $dbw->select( 'revision', array( 'rev_id' ), array( 'rev_page' => $pageID ), __METHOD__ );
 153                 $slaveIDs = array();
 154                 foreach ( $res as $row ) {
 155                         $slaveIDs[] = $row->rev_id;
 156                 }
 157                 if ( count( $masterIDs ) < count( $slaveIDs ) ) {
 158                         $missingIDs = array_diff( $slaveIDs, $masterIDs );
 159                         if ( count( $missingIDs ) ) {
 160                                 $this->output( "Found " . count( $missingIDs )
 161                                         . " lost in master, copying from slave... " );
 162                                 $dbFrom = $dbw;
 163                                 $found = true;
 164                                 $toMaster = true;
 165                         } else {
 166                                 $found = false;
 167                         }
 168                 } else {
 169                         $missingIDs = array_diff( $masterIDs, $slaveIDs );
 170                         if ( count( $missingIDs ) ) {
 171                                 $this->output( "Found " . count( $missingIDs )
 172                                         . " missing revision(s), copying from master... " );
 173                                 $dbFrom = $dbw;
 174                                 $found = true;
 175                                 $toMaster = false;
 176                         } else {
 177                                 $found = false;
 178                         }
 179                 }
 180
 181                 if ( $found ) {
 182                         foreach ( $missingIDs as $rid ) {
 183                                 $this->output( "$rid " );
 184                                 # Revision
 185                                 $row = $dbFrom->selectRow( 'revision', '*', array( 'rev_id' => $rid ), __METHOD__ );
 186                                 if ( $toMaster ) {
 187                                         $id = $dbw->selectField( 'revision', 'rev_id', array( 'rev_id' => $rid ),
 188                                                 __METHOD__, 'FOR UPDATE' );
 189                                         if ( $id ) {
 190                                                 $this->output( "Revision already exists\n" );
 191                                                 $found = false;
 192                                                 break;
 193                                         } else {
 194                                                 $dbw->insert( 'revision', get_object_vars( $row ), __METHOD__, 'IGNORE' );
 195                                         }
 196                                 } else {
 197                                         foreach ( $this->slaveIndexes as $i ) {
 198                                                 $db = wfGetDB( $i );
 199                                                 $db->insert( 'revision', get_object_vars( $row ), __METHOD__, 'IGNORE' );
 200                                         }
 201                                 }
 202
 203                                 # Text
 204                                 $row = $dbFrom->selectRow( 'text', '*', array( 'old_id' => $row->rev_text_id ), __METHOD__ );
 205                                 if ( $toMaster ) {
 206                                         $dbw->insert( 'text', get_object_vars( $row ), __METHOD__, 'IGNORE' );
 207                                 } else {
 208                                         foreach ( $this->slaveIndexes as $i ) {
 209                                                 $db = wfGetDB( $i );
 210                                                 $db->insert( 'text', get_object_vars( $row ), __METHOD__, 'IGNORE' );
 211                                         }
 212                                 }
 213                         }
 214                         $this->output( "done\n" );
 215                 }
 216
 217                 if ( $found ) {
 218                         $this->output( "Fixing page_latest... " );
 219                         if ( $toMaster ) {
 220                                 /*
 221                                 $dbw->update(
 222                                         'page',
 223                                         array( 'page_latest' => $realLatest ),
 224                                         array( 'page_id' => $pageID ),
 225                                         __METHOD__
 226                                 );
 227                                 */
 228                         } else {
 229                                 foreach ( $this->slaveIndexes as $i ) {
 230                                         $db = wfGetDB( $i );
 231                                         $db->update(
 232                                                 'page',
 233                                                 array( 'page_latest' => $realLatest ),
 234                                                 array( 'page_id' => $pageID ),
 235                                                 __METHOD__
 236                                         );
 237                                 }
 238                         }
 239                         $this->output( "done\n" );
 240                 }
 241                 $dbw->commit( __METHOD__ );
 242         }
 243 }
 244
 245 $maintClass = "FixSlaveDesync";
 246 require_once RUN_MAINTENANCE_IF_MAIN;