misc/cronjobs/cleanup_database.pl

   1 #!/usr/bin/perl
   2
   3 # Copyright 2009 PTFS, Inc.
   4 #
   5 # This file is part of Koha.
   6 #
   7 # Koha is free software; you can redistribute it and/or modify it under the
   8 # terms of the GNU General Public License as published by the Free Software
   9 # Foundation; either version 2 of the License, or (at your option) any later
  10 # version.
  11 #
  12 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  13 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  14 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  15 #
  16 # You should have received a copy of the GNU General Public License along
  17 # with Koha; if not, write to the Free Software Foundation, Inc.,
  18 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  19
  20 use strict;
  21 use warnings;
  22
  23 use constant DEFAULT_ZEBRAQ_PURGEDAYS => 30;
  24 use constant DEFAULT_MAIL_PURGEDAYS => 30;
  25 use constant DEFAULT_IMPORT_PURGEDAYS => 60;
  26 use constant DEFAULT_LOGS_PURGEDAYS => 180;
  27 use constant DEFAULT_SEARCHHISTORY_PURGEDAYS => 30;
  28
  29 BEGIN {
  30     # find Koha's Perl modules
  31     # test carefully before changing this
  32     use FindBin;
  33     eval { require "$FindBin::Bin/../kohalib.pl" };
  34 }
  35
  36 use C4::Context;
  37 use C4::Dates;
  38
  39 use C4::Search;
  40
  41 use Getopt::Long;
  42
  43 sub usage {
  44     print STDERR <<USAGE;
  45 Usage: $0 [-h|--help] [--sessions] [--sessdays DAYS] [-v|--verbose] [--zebraqueue DAYS] [-m|--mail] [--merged] [--import DAYS] [--logs DAYS] [--searchhistory DAYS]
  46
  47    -h --help          prints this help message, and exits, ignoring all
  48                       other options
  49    --sessions         purge the sessions table.  If you use this while users
  50                       are logged into Koha, they will have to reconnect.
  51    --sessdays DAYS    purge only sessions older than DAYS days.
  52    -v --verbose       will cause the script to give you a bit more information
  53                       about the run.
  54    --zebraqueue DAYS  purge completed zebraqueue entries older than DAYS days.
  55                       Defaults to 30 days if no days specified.
  56    -m --mail DAYS     purge items from the mail queue that are older than DAYS days.
  57                       Defaults to 30 days if no days specified.
  58    --merged           purged completed entries from need_merge_authorities.
  59    --import DAYS      purge records from import tables older than DAYS days.
  60                       Defaults to 60 days if no days specified.
  61    --logs DAYS        purge entries from action_logs older than DAYS days.
  62                       Defaults to 180 days if no days specified.
  63    --searchhistory DAYS  purge entries from search_history older than DAYS days.
  64                          Defaults to 30 days if no days specified
  65 USAGE
  66     exit $_[0];
  67 }
  68
  69 my ( $help, $sessions, $sess_days, $verbose, $zebraqueue_days, $mail, $purge_merged, $pImport, $pLogs, $pSearchhistory);
  70
  71 GetOptions(
  72     'h|help'       => \$help,
  73     'sessions'     => \$sessions,
  74     'sessdays:i'   => \$sess_days,
  75     'v|verbose'    => \$verbose,
  76     'm|mail:i'       => \$mail,
  77     'zebraqueue:i' => \$zebraqueue_days,
  78     'merged'       => \$purge_merged,
  79     'import:i'     => \$pImport,
  80     'logs:i'       => \$pLogs,
  81     'searchhistory:i' => \$pSearchhistory,
  82 ) || usage(1);
  83
  84 $sessions=1 if $sess_days && $sess_days>0;
  85 # if --import, --logs, --zebraqueue or --searchhistory were passed without number of days,
  86 # use defaults
  87 $pImport= DEFAULT_IMPORT_PURGEDAYS if defined($pImport) && $pImport==0;
  88 $pLogs= DEFAULT_LOGS_PURGEDAYS if defined($pLogs) && $pLogs==0;
  89 $zebraqueue_days= DEFAULT_ZEBRAQ_PURGEDAYS if defined($zebraqueue_days) && $zebraqueue_days==0;
  90 $mail= DEFAULT_MAIL_PURGEDAYS if defined($mail) && $mail==0;
  91 $pSearchhistory= DEFAULT_SEARCHHISTORY_PURGEDAYS if defined($pSearchhistory) && $pSearchhistory==0;
  92
  93 if ($help) {
  94     usage(0);
  95 }
  96
  97 if ( !( $sessions || $zebraqueue_days || $mail || $purge_merged || $pImport || $pLogs || $pSearchhistory ) ) {
  98     print "You did not specify any cleanup work for the script to do.\n\n";
  99     usage(1);
 100 }
 101
 102 my $dbh = C4::Context->dbh();
 103 my $query;
 104 my $sth;
 105 my $sth2;
 106 my $count;
 107
 108 if ( $sessions && !$sess_days ) {
 109     if ($verbose) {
 110         print "Session purge triggered.\n";
 111         $sth = $dbh->prepare("SELECT COUNT(*) FROM sessions");
 112         $sth->execute() or die $dbh->errstr;
 113         my @count_arr = $sth->fetchrow_array;
 114         print "$count_arr[0] entries will be deleted.\n";
 115     }
 116     $sth = $dbh->prepare("TRUNCATE sessions");
 117     $sth->execute() or die $dbh->errstr;
 118     if ($verbose) {
 119         print "Done with session purge.\n";
 120     }
 121 } elsif ( $sessions && $sess_days > 0 ) {
 122     if ($verbose) {
 123         print "Session purge triggered with days>$sess_days.\n";
 124     }
 125     RemoveOldSessions();
 126     if ($verbose) {
 127         print "Done with session purge with days>$sess_days.\n";
 128     }
 129 }
 130
 131 if ($zebraqueue_days) {
 132     $count = 0;
 133     if ($verbose) {
 134         print "Zebraqueue purge triggered for $zebraqueue_days days.\n";
 135     }
 136     $sth = $dbh->prepare(
 137         "SELECT id,biblio_auth_number,server,time FROM zebraqueue
 138                           WHERE done=1 and time < date_sub(curdate(), interval ? day)"
 139     );
 140     $sth->execute($zebraqueue_days) or die $dbh->errstr;
 141     $sth2 = $dbh->prepare("DELETE FROM zebraqueue WHERE id=?");
 142     while ( my $record = $sth->fetchrow_hashref ) {
 143         $sth2->execute( $record->{id} ) or die $dbh->errstr;
 144         $count++;
 145     }
 146     if ($verbose) {
 147         print "$count records were deleted.\nDone with zebraqueue purge.\n";
 148     }
 149 }
 150
 151 if ($mail) {
 152     print "Mail queue purge triggered for $mail days.\n" if ($verbose);
 153
 154     $sth = $dbh->prepare("DELETE FROM message_queue WHERE time_queued < date_sub(curdate(), interval ? day)");
 155     $sth->execute($mail) or die $dbh->errstr;
 156     my $count = $sth->rows;
 157     $sth->finish;
 158
 159     print "$count messages were deleted from the mail queue.\nDone with message_queue purge.\n" if ($verbose);
 160 }
 161
 162 if($purge_merged) {
 163     print "Purging completed entries from need_merge_authorities.\n" if $verbose;
 164     $sth = $dbh->prepare("DELETE FROM need_merge_authorities WHERE done=1");
 165     $sth->execute() or die $dbh->errstr;
 166     print "Done with purging need_merge_authorities.\n" if $verbose;
 167 }
 168
 169 if($pImport) {
 170     print "Purging records from import tables.\n" if $verbose;
 171     PurgeImportTables();
 172     print "Done with purging import tables.\n" if $verbose;
 173 }
 174
 175 if($pLogs) {
 176     print "Purging records from action_logs.\n" if $verbose;
 177     $sth = $dbh->prepare("DELETE FROM action_logs WHERE timestamp < date_sub(curdate(), interval ? DAY)");
 178     $sth->execute($pLogs) or die $dbh->errstr;
 179     print "Done with purging action_logs.\n" if $verbose;
 180 }
 181
 182 if($pSearchhistory) {
 183     print "Purging records older than $pSearchhistory from search_history.\n" if $verbose;
 184     PurgeSearchHistory($pSearchhistory);
 185     print "Done with purging search_history.\n" if $verbose;
 186 }
 187
 188 exit(0);
 189
 190 sub RemoveOldSessions {
 191     my ( $id, $a_session, $limit, $lasttime );
 192     $limit = time() - 24 * 3600 * $sess_days;
 193
 194     $sth = $dbh->prepare("SELECT id, a_session FROM sessions");
 195     $sth->execute or die $dbh->errstr;
 196     $sth->bind_columns( \$id, \$a_session );
 197     $sth2  = $dbh->prepare("DELETE FROM sessions WHERE id=?");
 198     $count = 0;
 199
 200     while ( $sth->fetch ) {
 201         $lasttime = 0;
 202         if ( $a_session =~ /lasttime:\s+'?(\d+)/ ) {
 203             $lasttime = $1;
 204         } elsif ( $a_session =~ /(ATIME|CTIME):\s+'?(\d+)/ ) {
 205             $lasttime = $2;
 206         }
 207         if ( $lasttime && $lasttime < $limit ) {
 208             $sth2->execute($id) or die $dbh->errstr;
 209             $count++;
 210         }
 211     }
 212     if ($verbose) {
 213         print "$count sessions were deleted.\n";
 214     }
 215 }
 216
 217 sub PurgeImportTables {
 218     #First purge import_records
 219     #Delete cascades to import_biblios, import_items and import_record_matches
 220     $sth = $dbh->prepare("DELETE FROM import_records WHERE upload_timestamp < date_sub(curdate(), interval ? DAY)");
 221     $sth->execute($pImport) or die $dbh->errstr;
 222
 223     # Now purge import_batches
 224     # Timestamp cannot be used here without care, because records are added
 225     # continuously to batches without updating timestamp (z3950 search).
 226     # So we only delete older empty batches.
 227     # This delete will therefore not have a cascading effect.
 228     $sth = $dbh->prepare("DELETE ba
 229  FROM import_batches ba
 230  LEFT JOIN import_records re ON re.import_batch_id=ba.import_batch_id
 231  WHERE re.import_record_id IS NULL AND
 232  ba.upload_timestamp < date_sub(curdate(), interval ? DAY)");
 233     $sth->execute($pImport) or die $dbh->errstr;
 234 }