Hi
I'm curious if anyone is successfully either:
1) using AWS OpenSearch with Archivematica, instead of ElasticSearch?
2) using ElasticSearch v7.x with Archivematica, instead of v6.x?
OpenSearch is Amazon's cloud based fork of open source Elasticsearch 7.10, and according to the AWS documentation the REST APIs for ingest, search, and management are backwards compatible + query syntax and responses are also the same. They do note though that despite this, "some clients or tools may include code, such as version checks, that may cause the client or tool to not work with OpenSearch."
According the latest Archivematica documentation, AM v1.13 requires ElasticSearch version 6.x (and is tested with 6.5.4), and digging around in the issues on GitHub and other posts in the group I get the impression there may be some issues if you do try using ElasticSearch v7.x
We're just experimenting with using OpenSearch 1.1 now, and while AM seems to be able to index items OK, search queries are failing with errors such as:
RequestError: RequestError(400, u'search_phase_execution_exception', u'Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [sipuuid] in order to load field data by uninverting the inverted index. Note that this can use significant memory.')
We're continuing to investigate, but I'm conscious that if AM doesn't even support ElasticSearch 7.x, it isn't going to support OpenSearch, so using it may not currently be an option at all, so I'd love to hear if anyone else has tried this!
Thanks
John Gostick