From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Philip Semanchuk <philip(at)americanefficient(dot)com> |
Cc: | Vijaykumar Jain <vijaykumarjain(dot)github(at)gmail(dot)com>, PostgreSQL General <pgsql-general(at)lists(dot)postgresql(dot)org> |
Subject: | Re: CREATE/REFRESH MATERIALIZED VIEW planner difference? |
Date: | 2021-06-01 18:51:05 |
Message-ID: | 2847072.1622573465@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Philip Semanchuk <philip(at)americanefficient(dot)com> writes:
> I can confirm that it’s not waiting on a lock. In addition, through the AWS CPU utilization monitor I can see that the REFRESH uses one CPU/worker whereas the CREATE uses four. This is consistent with the EXPLAIN ANALYZE for the CREATE which says it uses four workers.
Hm. I tried to reproduce this here, and in a simple test case I get
parallelized plans for both CREATE and REFRESH. Are you sure the
REFRESH is running with the same server parameter settings?
>> also, can you share the plans where you see the diff.
> Unless I misunderstand, there is no plan for a REFRESH.
EXPLAIN isn't bright about that, but if you enable auto_explain,
it will log the plan for a REFRESH's query.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Vijaykumar Jain | 2021-06-01 19:15:25 | Re: CREATE/REFRESH MATERIALIZED VIEW planner difference? |
Previous Message | Philip Semanchuk | 2021-06-01 18:26:00 | Re: CREATE/REFRESH MATERIALIZED VIEW planner difference? |