I'm almost certain that this is the correct query, I know because it took an entire night to run and I saved a copy of what was on my browser before going to sleep in case it timed out.
What's making me lead towards a problem other than the query is that only ~half the documents got updated, and as far as I can tell data up to ~6 months ago had the query run on it but didn't lose any data.
I should be able to rebuild the field from a different source, but it might take close to forever.
Ok, I believe I'm at fault here. There is no bug, just an unattentive user. The data from July onwards was corrupted because I was uploading it to elastic as such, and everything functions as intended.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.