There is a set of data that I want to fit into Elasticsearch. Product description - a few paragraphs. I have a lot of them - about 250mln.
At the same time I want to remove stop words, hunspell and a couple of other things.
But I also want to leave the ability to search for text without analyzer, just default Text (well, maybe lead to a single case).
Because I suspect that in some cases, hunspell or filter may remove what you need or tell and just by exact match will not be able to find.
Questions:
-What is the best solution to index the same text using anaylzer and Default text I see - that you can do just two fields. With anaylzer and text. But this will increase memory usage by half, won't it? And I'm afraid it could also double hard disk consumption as well. (the same paragraph will be saved in its original form twice for each of the two fields)
-How do I search? Well, you can make two requests, one at a time. First you need to query first one and then the other. (Text and analyzer). Are there any special requests that aggregate information in one response? For example with priority.