Thread

  1. COPY FROM - to avoid WAL generation

    Ravi Krishna <sravikrishna@aol.com> — 2018-08-21T15:00:03Z

    In a recent thread of mine I learned something very interesting.  If a table is created and data is loaded via COPY FROM within the same transaction, then PG will be smart enough to not generate WAL logs because all it needs to do is to track the status of the transaction and let the data load go to the new data file created for the table.  If committed, the table is released for other sessions, if rolledback, vaccum will delete the data file later on.
    I tested it as follows for a table with 50 milllion rows.  No indexes.
    Case 1  - create the table first.  - in a separate transaction load the 50 million rows.
    Took 3 min 22 seconds
    Case 2  - start transaction  - create table  - load 50 million rows  - commit transaction
    Took: 3 min 16 seconds.
    Am I missing anything?
    
  2. Re: COPY FROM - to avoid WAL generation

    Jeff Ross <jross@openvistas.net> — 2018-08-21T15:26:11Z

    On 8/21/18 9:00 AM, Ravi Krishna wrote:
    
    > In a recent thread of mine I learned something very interesting. If a 
    > table is created and data is loaded via COPY FROM within the same 
    > transaction, then PG will be smart enough to not generate WAL logs 
    > because all it needs to do is to track the status of the transaction 
    > and let the data load go to the new data file created for the table.  
    > If committed, the table is released for other sessions, if rolledback, 
    > vaccum will delete the data file later on.
    >
    > I tested it as follows for a table with 50 milllion rows. No indexes.
    >
    > Case 1
    >   - create the table first.
    >   - in a separate transaction load the 50 million rows.
    >
    > Took 3 min 22 seconds
    >
    > Case 2
    >   - start transaction
    >   - create table
    >   - load 50 million rows
    >   - commit transaction
    >
    > Took: 3 min 16 seconds.
    >
    > Am I missing anything?
    >
    Have you looked into pg_bulkload?
    
    https://github.com/ossc-db/pg_bulkload
    
    Docs are here:
    
    http://ossc-db.github.io/pg_bulkload/index.html
    
    Jeff
    
  3. Re: COPY FROM - to avoid WAL generation

    Andres Freund <andres@anarazel.de> — 2018-08-21T15:31:28Z

    Hi,
    
    On 2018-08-21 15:00:03 +0000, Ravi Krishna wrote:
    > In a recent thread of mine I learned something very interesting.  If a table is created and data is loaded via COPY FROM within the same transaction, then PG will be smart enough to not generate WAL logs because all it needs to do is to track the status of the transaction and let the data load go to the new data file created for the table.  If committed, the table is released for other sessions, if rolledback, vaccum will delete the data file later on.
    > I tested it as follows for a table with 50 milllion rows.  No indexes.
    
    Please note this is only the case if wal_level = minimal. If replication
    (or PITR) is supported, that mode can't be used, because the data has to
    go into the WAL.
    
    Were you using wal_level = minimal?
    
    (FWIW, it's not VACUUM that'd unlink the data in cause of failure, but
    that doesn't really matter much).
    
    Greetings,
    
    Andres Freund