If i already have a cluster key1 for existing table, i want to change cluster key to key2 using
ALTER TABLE table CLUSTER BY (key2), then run OPTIMIZE table, based on databrick document , existing files will not be rewritten (verified by my test as well), cluster Key2 will only be applied incrementally. If my existing files are huge (saying having long history), using Key2 will not improve performance when i search the table (e.g. select * from table where key2='xxxx'). is liquid cluster supposed to work this way or i am missing anything?
