It's not convenient if the users want to partition a table with string columns, because the range can't be certain in most of cases, for example, try to partition the table by idc and host:
CREATE TABLE monitor(
idc STRING,
host STRING,
cpu_util DOUBLE,
disk_util DOUBLE,
ts TIMESTAMP TIME INDEX)
engine=mito PARTITION ON COLUMNS(idc, host) (
....how to write the expression?...
);
Perhaps a hash partition method would be more suitable in this scenario, just like:
PARTITION ON COLUMNS(idc, host) (
hash(idc, host) % 10
);
Partition the table into 10 regions by the value of hash(idc, host) % 10.
What type of enhancement is this?
API improvement, User experience
What does the enhancement do?
It's not convenient if the users want to partition a table with string columns, because the range can't be certain in most of cases, for example, try to partition the table by
idc
andhost
:Perhaps a hash partition method would be more suitable in this scenario, just like:
Partition the table into 10 regions by the value of
hash(idc, host) % 10
.Implementation challenges
No response