| From: | Jeff Davis <pgsql(at)j-davis(dot)com> |
|---|---|
| To: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Rangejoin rebased |
| Date: | 2018-01-17 05:49:15 |
| Message-ID: | CAMp0ubdH_5BpW60O6YkStJsjUaE+LL330McG8p3uffMg0MtRLQ@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Wed, Jan 10, 2018 at 7:49 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> Do we optimize for TIMESTAMP <@ RANGE as well?
Not currently. It requires a little extra complexity because empty
ranges will match anything and need special handling.
> Does this link in nicely with partition-aware joins?
Yes: if the partitioning is on a non-range column, and the join key
includes both the partition key and a range column, it can do
partition-wise joins.
It does not try to invent a concept of partitioning on a spatial key.
> Does it allow partition exclusion if you join a daterange to a time
> range partitioned table?
I'm a little unclear what you mean here. Are you talking about spatial
partitioning? Or are you talking about joining a daterange column to a
timestamptz column (I suppose using @>)? I think the answer to your
question is "no", but let me know if I am missing an important case.
Regards,
Jeff Davis
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Amit Langote | 2018-01-17 06:03:24 | Re: TOAST table created for partitioned tables |
| Previous Message | Amit Langote | 2018-01-17 05:34:52 | Re: pgsql: Centralize json and jsonb handling of datetime types |