3 # This compares record counts from a Koha database to Elasticsearch
5 # Copyright 2019 ByWater Solutions
7 # This file is part of Koha.
9 # Koha is free software; you can redistribute it and/or modify it under the
10 # terms of the GNU General Public License as published by the Free Software
11 # Foundation; either version 3 of the License, or (at your option) any later
14 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
15 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
16 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
18 # You should have received a copy of the GNU General Public License along
19 # with Koha; if not, write to the Free Software Foundation, Inc.,
20 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 compare_es_to_db.pl - compares record counts from a Koha database to Elasticsearch
28 B<compare_es_to_db.pl>
33 use Array::Utils qw( array_diff );
37 use Koha::Authorities;
40 use Koha::SearchEngine::Elasticsearch;
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');
47 my $searcher = Koha::SearchEngine::Elasticsearch->new({ index => $index });
48 my $es = $searcher->get_elasticsearch();
49 my $count = $es->indices->stats( index => $searcher->get_elasticsearch_params->{index_name} )
50 ->{_all}{primaries}{docs}{count};
51 print "Count in db for $index is " . scalar @db_records . ", count in Elasticsearch is $count\n";
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->get_elasticsearch_params->{index_name},
69 # Here is where we actually iterate through
70 # Fetching each record, pushing the id into the array
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};
79 print "\nComparing arrays, this may take a while\n";
81 # And compare the arrays
82 # array_diff returns only a list of values which are not common to both arrays
83 my @diff = array_diff(@db_records, @es_ids );
85 print "All records match\n" unless @diff;
87 # Fetch values for providing record links
88 my $es_params = $searcher->get_elasticsearch_params;
89 my $es_base = "$es_params->{nodes}[0]/$es_params->{index_name}";
90 my $opac_base = C4::Context->preference('OPACBaseURL');
95 foreach my $problem ( sort { $a <=> $b } @diff){
96 if ( (grep /^$problem$/, @db_records) ){
97 push @koha_problems, $problem;
99 push @es_problems, $problem;
103 if ( @koha_problems ){
104 print "=================\n";
105 print "Records that exist in Koha but not in ES\n";
106 for my $problem ( @koha_problems ){
107 if ( $index eq 'biblios' ) {
109 print " Visit here to see record: $opac_base/cgi-bin/koha/opac-detail.pl?biblionumber=$problem\n";
110 } elsif ( $index eq 'authorities' ) {
112 print " Visit here to see record: $opac_base/cgi-bin/koha/opac-authoritiesdetail.pl?authid=$problem\n";
117 print "=================\n";
118 print "Records that exist in ES but not in Koha\n";
119 for my $problem ( @es_problems ){
121 print " Enter this command to view record: curl $es_base/data/$problem?pretty=true\n";