misc/maintenance/compare_es_to_db.pl

   1 #! /usr/bin/perl
   2 #
   3 # This compares record counts from a Koha database to Elasticsearch
   4
   5 # Copyright 2019 ByWater Solutions
   6 #
   7 # This file is part of Koha.
   8 #
   9 # Koha is free software; you can redistribute it and/or modify it
  10 # under the terms of the GNU General Public License as published by
  11 # the Free Software Foundation; either version 3 of the License, or
  12 # (at your option) any later version.
  13 #
  14 # Koha is distributed in the hope that it will be useful, but
  15 # WITHOUT ANY WARRANTY; without even the implied warranty of
  16 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17 # GNU General Public License for more details.
  18 #
  19 # You should have received a copy of the GNU General Public License
  20 # along with Koha; if not, see <http://www.gnu.org/licenses>.
  21
  22 =head1 NAME
  23
  24 compare_es_to_db.pl - compares record counts from a Koha database to Elasticsearch
  25
  26 =head1 SYNOPSIS
  27
  28 B<compare_es_to_db.pl>
  29
  30 =cut
  31
  32 use Modern::Perl;
  33 use Array::Utils qw( array_minus );
  34
  35 use C4::Context;
  36
  37 use Koha::Authorities;
  38 use Koha::Biblios;
  39 use Koha::Items;
  40 use Koha::SearchEngine::Elasticsearch;
  41
  42 foreach my $index ( ('biblios','authorities') ){
  43     print "=================\n";
  44     print "Checking $index\n";
  45     my @db_records = $index eq 'biblios' ? Koha::Biblios->search()->get_column('biblionumber') : Koha::Authorities->search()->get_column('authid');
  46
  47     my $searcher = Koha::SearchEngine::Elasticsearch->new({ index => $index });
  48     my $es = $searcher->get_elasticsearch();
  49     my $count = $es->indices->stats( index => $searcher->index_name )
  50         ->{_all}{primaries}{docs}{count};
  51     print "Count in db for $index is " . scalar @db_records . ", count in Elasticsearch is $count\n";
  52
  53     # Now we get all the ids from Elasticsearch
  54     # The scroll lets us iterate through, it fetches chunks of 'size' as we move through
  55     my $scroll = $es->scroll_helper(
  56         index => $searcher->index_name,
  57         size => 5000,
  58         body => {
  59             query => {
  60                 match_all => {}
  61             },
  62             stored_fields => []
  63         },
  64         scroll_in_qs => 1,
  65     );
  66
  67     my @es_ids;
  68
  69     # Here is where we actually iterate through
  70     # Fetching each record, pushing the id into the array
  71     my $i = 1;
  72     print "Fetching Elasticsearch records ids";
  73     while (my $doc = $scroll->next ){
  74         print "." if !($i % 500);
  75         print "\n$i records retrieved" if !($i % 5000);
  76         push @es_ids, $doc->{_id};
  77         $i++;
  78     }
  79
  80     # Fetch values for providing record links
  81     my $es_params = $searcher->get_elasticsearch_params;
  82     my $es_base   = "$es_params->{nodes}[0]/".$searcher->index_name;
  83     my $opac_base = C4::Context->preference('OPACBaseURL');
  84
  85     print "\nComparing arrays, this may take a while\n";
  86
  87     my @koha_problems = sort { $a <=> $b } array_minus(@db_records, @es_ids);
  88     my @es_problems = sort { $a <=> $b } array_minus(@es_ids, @db_records);
  89
  90     print "All records match\n" unless ( @koha_problems || @es_problems );
  91
  92     if ( @koha_problems ){
  93         print "=================\n";
  94         print "Records that exist in Koha but not in ES\n";
  95         for my $problem ( @koha_problems ){
  96             if ( $index eq 'biblios' ) {
  97                 print "  #$problem";
  98                 print "  Visit here to see record: $opac_base/cgi-bin/koha/opac-detail.pl?biblionumber=$problem\n";
  99             } elsif ( $index eq 'authorities' ) {
 100                 print "#$problem";
 101                 print "  Visit here to see record: $opac_base/cgi-bin/koha/opac-authoritiesdetail.pl?authid=$problem\n";
 102             }
 103         }
 104     }
 105
 106     if ( @es_problems ){
 107         print "=================\n";
 108         print "Records that exist in ES but not in Koha\n";
 109         for my $problem ( @es_problems ){
 110             print "  #$problem";
 111             print "  Enter this command to view record: curl $es_base/data/$problem?pretty=true\n";
 112         }
 113     }
 114 }