includes/filebackend/FileOpBatch.php

   1 <?php
   2 /**
   3  * Helper class for representing batch file operations.
   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 FileBackend
  22  * @author Aaron Schulz
  23  */
  24
  25 /**
  26  * Helper class for representing batch file operations.
  27  * Do not use this class from places outside FileBackend.
  28  *
  29  * Methods should avoid throwing exceptions at all costs.
  30  *
  31  * @ingroup FileBackend
  32  * @since 1.20
  33  */
  34 class FileOpBatch {
  35         /* Timeout related parameters */
  36         const MAX_BATCH_SIZE = 1000; // integer
  37
  38         /**
  39          * Attempt to perform a series of file operations.
  40          * Callers are responsible for handling file locking.
  41          *
  42          * $opts is an array of options, including:
  43          *   - force        : Errors that would normally cause a rollback do not.
  44          *                    The remaining operations are still attempted if any fail.
  45          *   - nonJournaled : Don't log this operation batch in the file journal.
  46          *   - concurrency  : Try to do this many operations in parallel when possible.
  47          *
  48          * The resulting Status will be "OK" unless:
  49          *   - a) unexpected operation errors occurred (network partitions, disk full...)
  50          *   - b) significant operation errors occurred and 'force' was not set
  51          *
  52          * @param $performOps Array List of FileOp operations
  53          * @param $opts Array Batch operation options
  54          * @param $journal FileJournal Journal to log operations to
  55          * @return Status
  56          */
  57         public static function attempt( array $performOps, array $opts, FileJournal $journal ) {
  58                 wfProfileIn( __METHOD__ );
  59                 $status = Status::newGood();
  60
  61                 $n = count( $performOps );
  62                 if ( $n > self::MAX_BATCH_SIZE ) {
  63                         $status->fatal( 'backend-fail-batchsize', $n, self::MAX_BATCH_SIZE );
  64                         wfProfileOut( __METHOD__ );
  65                         return $status;
  66                 }
  67
  68                 $batchId = $journal->getTimestampedUUID();
  69                 $ignoreErrors = !empty( $opts['force'] );
  70                 $journaled = empty( $opts['nonJournaled'] );
  71                 $maxConcurrency = isset( $opts['concurrency'] ) ? $opts['concurrency'] : 1;
  72
  73                 $entries = array(); // file journal entry list
  74                 $predicates = FileOp::newPredicates(); // account for previous ops in prechecks
  75                 $curBatch = array(); // concurrent FileOp sub-batch accumulation
  76                 $curBatchDeps = FileOp::newDependencies(); // paths used in FileOp sub-batch
  77                 $pPerformOps = array(); // ordered list of concurrent FileOp sub-batches
  78                 $lastBackend = null; // last op backend name
  79                 // Do pre-checks for each operation; abort on failure...
  80                 foreach ( $performOps as $index => $fileOp ) {
  81                         $backendName = $fileOp->getBackend()->getName();
  82                         $fileOp->setBatchId( $batchId ); // transaction ID
  83                         // Decide if this op can be done concurrently within this sub-batch
  84                         // or if a new concurrent sub-batch must be started after this one...
  85                         if ( $fileOp->dependsOn( $curBatchDeps )
  86                                 || count( $curBatch ) >= $maxConcurrency
  87                                 || ( $backendName !== $lastBackend && count( $curBatch ) )
  88                         ) {
  89                                 $pPerformOps[] = $curBatch; // push this batch
  90                                 $curBatch = array(); // start a new sub-batch
  91                                 $curBatchDeps = FileOp::newDependencies();
  92                         }
  93                         $lastBackend = $backendName;
  94                         $curBatch[$index] = $fileOp; // keep index
  95                         // Update list of affected paths in this batch
  96                         $curBatchDeps = $fileOp->applyDependencies( $curBatchDeps );
  97                         // Simulate performing the operation...
  98                         $oldPredicates = $predicates;
  99                         $subStatus = $fileOp->precheck( $predicates ); // updates $predicates
 100                         $status->merge( $subStatus );
 101                         if ( $subStatus->isOK() ) {
 102                                 if ( $journaled ) { // journal log entries
 103                                         $entries = array_merge( $entries,
 104                                                 $fileOp->getJournalEntries( $oldPredicates, $predicates ) );
 105                                 }
 106                         } else { // operation failed?
 107                                 $status->success[$index] = false;
 108                                 ++$status->failCount;
 109                                 if ( !$ignoreErrors ) {
 110                                         wfProfileOut( __METHOD__ );
 111                                         return $status; // abort
 112                                 }
 113                         }
 114                 }
 115                 // Push the last sub-batch
 116                 if ( count( $curBatch ) ) {
 117                         $pPerformOps[] = $curBatch;
 118                 }
 119
 120                 // Log the operations in the file journal...
 121                 if ( count( $entries ) ) {
 122                         $subStatus = $journal->logChangeBatch( $entries, $batchId );
 123                         if ( !$subStatus->isOK() ) {
 124                                 wfProfileOut( __METHOD__ );
 125                                 return $subStatus; // abort
 126                         }
 127                 }
 128
 129                 if ( $ignoreErrors ) { // treat precheck() fatals as mere warnings
 130                         $status->setResult( true, $status->value );
 131                 }
 132
 133                 // Attempt each operation (in parallel if allowed and possible)...
 134                 self::runParallelBatches( $pPerformOps, $status );
 135
 136                 wfProfileOut( __METHOD__ );
 137                 return $status;
 138         }
 139
 140         /**
 141          * Attempt a list of file operations sub-batches in series.
 142          *
 143          * The operations *in* each sub-batch will be done in parallel.
 144          * The caller is responsible for making sure the operations
 145          * within any given sub-batch do not depend on each other.
 146          * This will abort remaining ops on failure.
 147          *
 148          * @param $pPerformOps Array
 149          * @param $status Status
 150          * @return bool Success
 151          */
 152         protected static function runParallelBatches( array $pPerformOps, Status $status ) {
 153                 $aborted = false; // set to true on unexpected errors
 154                 foreach ( $pPerformOps as $performOpsBatch ) {
 155                         if ( $aborted ) { // check batch op abort flag...
 156                                 // We can't continue (even with $ignoreErrors) as $predicates is wrong.
 157                                 // Log the remaining ops as failed for recovery...
 158                                 foreach ( $performOpsBatch as $i => $fileOp ) {
 159                                         $performOpsBatch[$i]->logFailure( 'attempt_aborted' );
 160                                 }
 161                                 continue;
 162                         }
 163                         $statuses = array();
 164                         $opHandles = array();
 165                         // Get the backend; all sub-batch ops belong to a single backend
 166                         $backend = reset( $performOpsBatch )->getBackend();
 167                         // Get the operation handles or actually do it if there is just one.
 168                         // If attemptAsync() returns a Status, it was either due to an error
 169                         // or the backend does not support async ops and did it synchronously.
 170                         foreach ( $performOpsBatch as $i => $fileOp ) {
 171                                 if ( !$fileOp->failed() ) { // failed => already has Status
 172                                         // If the batch is just one operation, it's faster to avoid
 173                                         // pipelining as that can involve creating new TCP connections.
 174                                         $subStatus = ( count( $performOpsBatch ) > 1 )
 175                                                 ? $fileOp->attemptAsync()
 176                                                 : $fileOp->attempt();
 177                                         if ( $subStatus->value instanceof FileBackendStoreOpHandle ) {
 178                                                 $opHandles[$i] = $subStatus->value; // deferred
 179                                         } else {
 180                                                 $statuses[$i] = $subStatus; // done already
 181                                         }
 182                                 }
 183                         }
 184                         // Try to do all the operations concurrently...
 185                         $statuses = $statuses + $backend->executeOpHandlesInternal( $opHandles );
 186                         // Marshall and merge all the responses (blocking)...
 187                         foreach ( $performOpsBatch as $i => $fileOp ) {
 188                                 if ( !$fileOp->failed() ) { // failed => already has Status
 189                                         $subStatus = $statuses[$i];
 190                                         $status->merge( $subStatus );
 191                                         if ( $subStatus->isOK() ) {
 192                                                 $status->success[$i] = true;
 193                                                 ++$status->successCount;
 194                                         } else {
 195                                                 $status->success[$i] = false;
 196                                                 ++$status->failCount;
 197                                                 $aborted = true; // set abort flag; we can't continue
 198                                         }
 199                                 }
 200                         }
 201                 }
 202                 return $status;
 203         }
 204 }