I have a table that looks like this
CREATE TABLE t (id INT, ts TIMESTAMP, /* 90 other small fields */, PRIMARY KEY (id, ts)) WITH CLUSTERING ORDER BY (ts DESC) AND compaction = {'class': 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy', 'max_threshold': '32', 'min_threshold': '4'}
The table is supposed to be queried as
SELECT * FROM t WHERE id = 1234 LIMIT 1
and
SELECT * FROM t WHERE id = 1234 AND ts < some_ts LIMIT 1
It has 1.5B rows and 200M unique ids and is populated with SSTableLoader from some source data. After the data is loaded into Cassandra it starts compacting it which takes quite some time. So two questions:
1. Is the data model correct for intended purpose - write once, read many, getting either the last version of the row or a version for certain timestamp?
2. Could something be done at CQLTableSSTableWriter side to reduce compactions? I have full control on how the data is written.
