Found when testing Bug 16939. To reproduce: Set SearchEngine syspref to Elasticsearch Go to Acquisitions and find a basket Add an order to the basket from an existing record Confirm error Message in error logs: neworderbiblio.pl: [NoNodes] ** No nodes are available: [http://localhost:9200], called from sub Search::Elasticsearch::Role::Client::Direct::__ANON__ at /usr/share/perl5/Catmandu/Store/ElasticSearch.pm line 27. at /home/vagrant/kohaclone/Koha/SearchEngine/Elasticsearch.pm line 437., referer: http://localhost:8081/cgi-bin/koha/acqui/booksellers.pl neworderbiblio.pl: Unable to perform your search. Please try again., referer: http://localhost:8081/cgi-bin/koha/acqui/booksellers.pl
The "No nodes are available" message suggests an issue with your ES server, can you confirm ES is running and searchign si working and then see if this can be recreated? I can add to baskets with no issues.
I am facing the same issue. After upgrading to the latest: $ koha-elasticsearch --verbose --rebuild $instance Indexing biblios [NoNodes] ** No nodes are available: [vpc-els-w2lu.ap-southeast-1.es.amazonaws.com], called from sub Search::Elasticsearch::Role::Client::Direct::__ANON__ at /usr/share/koha/lib/Koha/SearchEngine/Elasticsearch/Indexer.pm line 382. ES is up and running with 2 nodes. Curl response from Koha server: { "name" : "PiJKGkk", "cluster_name" : "18949800000:els", "cluster_uuid" : "KJB8jbkjhjki78KJKJ88", "version" : { "number" : "5.5.2", "build_hash" : "b2f0c09", "build_date" : "2018-09-11T13:05:43.663Z", "build_snapshot" : true, "lucene_version" : "6.6.0" }, "tagline" : "You Know, for Search" }
Koha Version 18.11.00.000
Same here, search is working ok, so it's not a configuration issue. I'd say the problem is that we are passing the result of build_query_compat to simple_search_compat. Regarding other calls it expects a string. It also appears from addbooks.pl The POD says both are ok: """ A thing to search for. It could be a simple string, or something constructed with the appropriate QueryBuilder module. """ Hard to say what is expected as this method is not covered by tests (!) Upping severity.
(In reply to Reza from comment #2) > I am facing the same issue. I do not think it is the same issue, it looks like a configuration problem.
Scratch the last comment, I get the same error with a "normal search"
(for the record) Works for me finally, I had to upgrade elastic and: % cd /usr/share/elasticsearch % sudo bin/elasticsearch-plugin remove analysis-icu % sudo bin/elasticsearch-plugin install analysis-icu hit /cgi-bin/koha/admin/searchengine/elasticsearch/mappings.pl?op=reset&i_know_what_i_am_doing=1 % perl misc/search_tools/rebuild_elastic_search.pl -d % sudo koha-elasticsearch --rebuild -b -a kohadev not trivial...
I contacted Catalyst to solve the issue for us. Our Koha is installed with a memcached and an elastic search cluster with two nodes. This is the finding: On line 92 of Koha::SearchEngine::Elasticsearch I removed cxn_pool => 'Sniff', This allowed us to connect to the Amazon Elastic cluster. I then dropped and recreated the Elasticsearch indexes I note that it errored if you tried to do the default 5000 records per commit, and that it still didn't like 1000, but it was happy at 100 So when you run koha-elasticsearch --rebuild apiit you want to also add -c 100 so it does 100 records at a time.
Reza, I have opened bug 21974 to deal with cxn_pool. I think the original error reported here was a configuration problem as Aleisha was testing it locally.