Remove cutoff_frequency, deprecated in es7.3.0, forbidden in es8 #1657
Conversation
michaelkirk
force-pushed
the
mkirk/remove-cutoff-freq
branch
from
9702be4 to
f2f641a
Compare
|
re: build failure https://github.com/pelias/api/actions/runs/5627436077 Because of fa9838f, it looks like I need to add my own repo/org variables in order for tests to pass. I've done that here: 
...and now tests are passing (except for the docker --push, for which I haven't configured secrets. It seems like it'd be more convenient to use the pre-existing It's not a big deal, but I expect others will get confused by this as well. |
|
Hi @michaelkirk, I've merged pelias/query#134, can you please update this PR so the
.. and then open it up for review, tx! |
From https://www.elastic.co/guide/en/elasticsearch/reference/8.8/migrating-8.0.html The cutoff_frequency parameter has been removed from the match and multi_match query. Details The cutoff_frequency parameter, deprecated in 7.x, has been removed in 8.0 from match and multi_match queries. The same functionality can be achieved without any configuration provided that the total number of hits is not tracked. Impact Discontinue use of the cutoff_frequency parameter. Search requests containing this parameter in a match or multi_match query will return an error. Note in the above "...provided that the total number of hits is not tracked". `track_total_hits` does not appear in the pelias codebases, so we shouldn't have any issues there.
michaelkirk
force-pushed
the
mkirk/remove-cutoff-freq
branch
from
f2f641a to
9282c62
Compare
|
Done (and rebased). |
|
I missed an occurrence somehow. Let me fix this up, retest, and I'll re-open shortly. |
|
You didn't miss the |
Here's the reason for this change 🚀
In pursuit of eventually supporting es8, I've dropped some es6 only behavior in pelias/query#134.
It's a draft because it depends on pelias/query#134 being released first.
Here's what actually got changed 👏
cutoff_frequency.Here's how others can test the changes 👀
Other than
npm test, I've run the north-america tests.Note: The north-america tests aren't all passing, but the ones that are failing are exactly the same as when I run from current master. I'm assuming (🤞) that the failures reflect changes in the input data since the tests were updated, as hypothesized over here.
north-america integration test output before
north-america test output after