Merge "Drop characters in Unicode Private Use Area in conversion tables."
[lhc/web/wiklou.git] / 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 * @ingroup Maintenance
21 */
22
23 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
24
25 /**
26 * Maintenance script that fixes erroneous page_latest values
27 * due to slave desynchronisation.
28 *
29 * @ingroup Maintenance
30 */
31 class FixSlaveDesync extends Maintenance {
32 public function __construct() {
33 parent::__construct();
34 $this->mDescription = "";
35 }
36
37 public function getDbType() {
38 return Maintenance::DB_ADMIN;
39 }
40
41 public function execute() {
42 $this->slaveIndexes = array();
43 for ( $i = 1; $i < wfGetLB()->getServerCount(); $i++ ) {
44 if ( wfGetLB()->isNonZeroLoad( $i ) ) {
45 $this->slaveIndexes[] = $i;
46 }
47 }
48
49 if ( $this->hasArg() ) {
50 $this->desyncFixPage( $this->getArg() );
51 } else {
52 $corrupt = $this->findPageLatestCorruption();
53 foreach ( $corrupt as $id => $dummy ) {
54 $this->desyncFixPage( $id );
55 }
56 }
57 }
58
59 /**
60 * Find all pages that have a corrupted page_latest
61 * @return array
62 */
63 private function findPageLatestCorruption() {
64 $desync = array();
65 $n = 0;
66 $dbw = wfGetDB( DB_MASTER );
67 $masterIDs = array();
68 $res = $dbw->select( 'page', array( 'page_id', 'page_latest' ), array( 'page_id<6054123' ), __METHOD__ );
69 $this->output( "Number of pages: " . $dbw->numRows( $res ) . "\n" );
70 foreach ( $res as $row ) {
71 $masterIDs[$row->page_id] = $row->page_latest;
72 if ( !( ++$n % 10000 ) ) {
73 $this->output( "$n\r" );
74 }
75 }
76 $this->output( "\n" );
77
78 foreach ( $this->slaveIndexes as $i ) {
79 $db = wfGetDB( $i );
80 $res = $db->select( 'page', array( 'page_id', 'page_latest' ), array( 'page_id<6054123' ), __METHOD__ );
81 foreach ( $res as $row ) {
82 if ( isset( $masterIDs[$row->page_id] ) && $masterIDs[$row->page_id] != $row->page_latest ) {
83 $desync[$row->page_id] = true;
84 $this->output( $row->page_id . "\t" );
85 }
86 }
87 }
88 $this->output( "\n" );
89 return $desync;
90 }
91
92 /**
93 * Fix a broken page entry
94 * @param $pageID int The page_id to fix
95 */
96 private function desyncFixPage( $pageID ) {
97 # Check for a corrupted page_latest
98 $dbw = wfGetDB( DB_MASTER );
99 $dbw->begin( __METHOD__ );
100 $realLatest = $dbw->selectField( 'page', 'page_latest', array( 'page_id' => $pageID ),
101 __METHOD__, 'FOR UPDATE' );
102 # list( $masterFile, $masterPos ) = $dbw->getMasterPos();
103 $found = false;
104 foreach ( $this->slaveIndexes as $i ) {
105 $db = wfGetDB( $i );
106 /*
107 if ( !$db->masterPosWait( $masterFile, $masterPos, 10 ) ) {
108 $this->output( "Slave is too lagged, aborting\n" );
109 $dbw->commit( __METHOD__ );
110 sleep(10);
111 return;
112 }*/
113 $latest = $db->selectField( 'page', 'page_latest', array( 'page_id' => $pageID ), __METHOD__ );
114 $max = $db->selectField( 'revision', 'MAX(rev_id)', false, __METHOD__ );
115 if ( $latest != $realLatest && $realLatest < $max ) {
116 $this->output( "page_latest corrupted in page $pageID, server $i\n" );
117 $found = true;
118 break;
119 }
120 }
121 if ( !$found ) {
122 $this->output( "page_id $pageID seems fine\n" );
123 $dbw->commit( __METHOD__ );
124 return;
125 }
126
127 # Find the missing revisions
128 $res = $dbw->select( 'revision', array( 'rev_id' ), array( 'rev_page' => $pageID ),
129 __METHOD__, 'FOR UPDATE' );
130 $masterIDs = array();
131 foreach ( $res as $row ) {
132 $masterIDs[] = $row->rev_id;
133 }
134
135 $res = $dbw->select( 'revision', array( 'rev_id' ), array( 'rev_page' => $pageID ), __METHOD__ );
136 $slaveIDs = array();
137 foreach ( $res as $row ) {
138 $slaveIDs[] = $row->rev_id;
139 }
140 if ( count( $masterIDs ) < count( $slaveIDs ) ) {
141 $missingIDs = array_diff( $slaveIDs, $masterIDs );
142 if ( count( $missingIDs ) ) {
143 $this->output( "Found " . count( $missingIDs ) . " lost in master, copying from slave... " );
144 $dbFrom = $dbw;
145 $found = true;
146 $toMaster = true;
147 } else {
148 $found = false;
149 }
150 } else {
151 $missingIDs = array_diff( $masterIDs, $slaveIDs );
152 if ( count( $missingIDs ) ) {
153 $this->output( "Found " . count( $missingIDs ) . " missing revision(s), copying from master... " );
154 $dbFrom = $dbw;
155 $found = true;
156 $toMaster = false;
157 } else {
158 $found = false;
159 }
160 }
161
162 if ( $found ) {
163 foreach ( $missingIDs as $rid ) {
164 $this->output( "$rid " );
165 # Revision
166 $row = $dbFrom->selectRow( 'revision', '*', array( 'rev_id' => $rid ), __METHOD__ );
167 if ( $toMaster ) {
168 $id = $dbw->selectField( 'revision', 'rev_id', array( 'rev_id' => $rid ),
169 __METHOD__, 'FOR UPDATE' );
170 if ( $id ) {
171 $this->output( "Revision already exists\n" );
172 $found = false;
173 break;
174 } else {
175 $dbw->insert( 'revision', get_object_vars( $row ), __METHOD__, 'IGNORE' );
176 }
177 } else {
178 foreach ( $this->slaveIndexes as $i ) {
179 $db = wfGetDB( $i );
180 $db->insert( 'revision', get_object_vars( $row ), __METHOD__, 'IGNORE' );
181 }
182 }
183
184 # Text
185 $row = $dbFrom->selectRow( 'text', '*', array( 'old_id' => $row->rev_text_id ), __METHOD__ );
186 if ( $toMaster ) {
187 $dbw->insert( 'text', get_object_vars( $row ), __METHOD__, 'IGNORE' );
188 } else {
189 foreach ( $this->slaveIndexes as $i ) {
190 $db = wfGetDB( $i );
191 $db->insert( 'text', get_object_vars( $row ), __METHOD__, 'IGNORE' );
192 }
193 }
194 }
195 $this->output( "done\n" );
196 }
197
198 if ( $found ) {
199 $this->output( "Fixing page_latest... " );
200 if ( $toMaster ) {
201 # $dbw->update( 'page', array( 'page_latest' => $realLatest ), array( 'page_id' => $pageID ), __METHOD__ );
202 } else {
203 foreach ( $this->slaveIndexes as $i ) {
204 $db = wfGetDB( $i );
205 $db->update( 'page', array( 'page_latest' => $realLatest ), array( 'page_id' => $pageID ), __METHOD__ );
206 }
207 }
208 $this->output( "done\n" );
209 }
210 $dbw->commit( __METHOD__ );
211 }
212 }
213
214 $maintClass = "FixSlaveDesync";
215 require_once( RUN_MAINTENANCE_IF_MAIN );