Phoenix queries seem to return incorrect results when defining tables with more salt buckets than regions. Is this a known problem or am I misunderstanding something?
I tested with phoenix 5.1.2 and hbase 2.1.4. Create a table with 4 salt buckets and 2 initial regions and add some values: create table JVB (ID BIGINT NOT NULL, NAME VARCHAR, CONSTRAINT pk PRIMARY KEY (ID)) SALT_BUCKETS = 4 SPLIT ON (5); upsert into JVB values (0, 'A'); upsert into JVB values (1, 'A'); upsert into JVB values (2, 'A'); upsert into JVB values (3, 'A'); upsert into JVB values (4, 'A'); upsert into JVB values (5, 'A'); upsert into JVB values (6, 'A'); upsert into JVB values (7, 'A'); upsert into JVB values (8, 'A'); upsert into JVB values (9, 'A'); Example queries that are incorrect: 0: jdbc:phoenix:> select * from JVB where ID < 4; +----+------+ | ID | NAME | +----+------+ | 3 | A | +----+------+ 1 row selected (0.011 seconds) 0: jdbc:phoenix:> select * from JVB where ID > 4; +----+------+ | ID | NAME | +----+------+ | 7 | A | | 2 | A | | 6 | A | | 1 | A | | 5 | A | | 9 | A | | 0 | A | | 4 | A | | 8 | A | +----+------+ 9 rows selected (0.012 seconds) Thanks, Jan
