filter a table on a cluster key efficiently

I have a snowflake table, with a clustering key with the format yyyy-mm-dd-hh, and I would like to filter it on all the values of a certain date.

I have written the where clause in those ways:

1. WHERE dt >= current_date() || '-00' AND dt <= current_date() || '-23' - took 0.7 seconds
2. WHERE dt IN (current_date() || '-00',current_date() || '-01',current_date() || '-02',current_date() || '-03',current_date() || '-04',current_date() || '-05',current_date() || '-06',current_date() || '-07',current_date() || '-08',current_date() || '-09',current_date() || '-10',current_date() || '-11',current_date() || '-12',current_date() || '-13',current_date() || '-14',current_date() || '-15',current_date() || '-16',current_date() || '-17',current_date() || '-18',current_date() || '-19',current_date() || '-20',current_date() || '-21',current_date() || '-22',current_date() || '-23') - took the same as #1
3. WHERE dt LIKE current_date() || '%' - took 2.9 seconds
4. WHERE left(dt, 10) = current_date - took 9.2 seconds

and I set ALTER SESSION SET USE_CACHED_RESULT = FALSE to make sure the result cache is not being used

my questions are:

1. is the a more efficient way to query than #1?
2. what type of queries can take advantage of the clustering? I would expect #3 to perform similar to #1, so I'm wandering what are the rules to write an efficient query

https://stackoverflow.com/questions/65937717/filter-a-table-on-a-cluster-key-efficiently January 28, 2021 at 09:07PM