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