Related: Jason Baumgartner has maintained a Reddit scraping pipeline for a few years now, and wrote up some notes about making it robust: https://pushshift.io
Example query which searches for 'f5bot' in the past day and correctly finds the corresponding posts on Reddit:
#standardSQL
SELECT title, subreddit, permalink
FROM `pushshift.rt_reddit.submissions`
WHERE created_utc > TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 1 DAY)
AND REGEXP_CONTAINS(LOWER(title), r'f5bot')