This is a known issue, and is caused by the time it takes the browser to parse the large JSON response that comes back from Elasticsearch. It's mentioned on this issue, but more in regards to the UX:
There is also this issue which proposes allowing users to hide the contents of known large fields:
There are actually 2 problems here. The slowness of the JSON parsing is further complicated by the slowness of the table rendering in Discover, as mentioned here (and this is likely what is causing the issue for you):
We have talked about using a streaming JSON parser to deal with large response bodies, but that is still in more of the consideration phase. I thought we had an issue open for dealing with large documents specifically, which I thought had some discussion about a streaming parser, but I can't seem to find it.
As for fixing the table rendering, that last issue is probably the one to follow and leave comments on.
Thank you very much for the links. Unfortunately none of these describes the problem we have.
"Show progress bar for big result sets": like I wrote, the result set is not that big. 6 documents, one with 4.8MB in size, the others a few kilobytes maybe. The whole result set isn't more than 6MB. On a local GBit lan this shouldn't be a problem.
"Loading huge documents": this is more like what I described, but the proposed solution (eliminating huge fields) is a no-go. The huge field is the field we need the most from the result set.
"Slow doc table rendering with large number of fields": this seems more like it, but I don't think that 20-21 fields are that much. When I get the time I'll make a index with only 1 doc, 1 field and make the one field really big, then I'll test if the same problem occurs.
I really would like to make a video, but the information in the docs sensitive, so I'm not allowed to.
Also, I thought about making a timeline graph (similar to the ones shown in the first and third link), but I can't even open the page in Chrome (copied the link from FF). I always get a popup, where I can "Kill" or "Wait". I have clicked "Wait" 5 times in a row, but the page still didn't load.