Thread

  1. Re: PITR, checkpoint, and local relations

    Vadim Mikheev <vmikheev@sectorbase.com> — 2002-08-02T23:50:31Z

    > So I think what will work then is pg_copy (hot backup) would:
    > 1) Issue an ALTER SYSTEM BEGIN BACKUP command which turns on 
    > atomic write,
    > checkpoints the database and disables further checkpoints (so 
    > wal files
    > won't be reused) until the backup is complete.
    > 2) Change ALTER SYSTEM BACKUP DATABASE TO <directory> read 
    > the database
    > directory to find which files it should backup rather than 
    > pg_class and for
    > each file just use system(cp...) to copy it to the backup directory.
    
    Did you consider saving backup on the client host (ie from where
    pg_copy started)?
    
    > 3) ALTER SYSTEM FINISH BACKUP does at it does now and backs 
    > up the pg_xlog
    > directory and renables database checkpointing.
    
    Well, wouldn't be single command ALTER SYSTEM BACKUP enough?
    What's the point to have 3 commands?
    
    (If all of this is already discussed then sorry - I'm not going
    to start new discussion).
    
    Vadim
    
    
  2. Re: PITR, checkpoint, and local relations

    Richard Tucker <richt@multera.com> — 2002-08-07T15:01:59Z

    
    > -----Original Message-----
    > From: Mikheev, Vadim [mailto:vmikheev@SECTORBASE.COM]
    > Sent: Friday, August 02, 2002 7:51 PM
    > To: 'richt@multera.com'; J. R. Nield
    > Cc: Tom Lane; Bruce Momjian; PostgreSQL Hacker
    > Subject: RE: [HACKERS] PITR, checkpoint, and local relations
    >
    >
    > > So I think what will work then is pg_copy (hot backup) would:
    > > 1) Issue an ALTER SYSTEM BEGIN BACKUP command which turns on
    > > atomic write,
    > > checkpoints the database and disables further checkpoints (so
    > > wal files
    > > won't be reused) until the backup is complete.
    > > 2) Change ALTER SYSTEM BACKUP DATABASE TO <directory> read
    > > the database
    > > directory to find which files it should backup rather than
    > > pg_class and for
    > > each file just use system(cp...) to copy it to the backup directory.
    >
    > Did you consider saving backup on the client host (ie from where
    > pg_copy started)?
    No, pg_copy just uses the libpq interface.
    >
    > > 3) ALTER SYSTEM FINISH BACKUP does at it does now and backs
    > > up the pg_xlog
    > > directory and renables database checkpointing.
    >
    I think now it could be just one command.  My implementation was reading
    pg_class to find the tables and indexes that needed backing up.  Now reading
    pg_database would be sufficient to find the directories containing files
    that needed to be archived, so it could all be done in one command.
    
    > Well, wouldn't be single command ALTER SYSTEM BACKUP enough?
    > What's the point to have 3 commands?
    >
    > (If all of this is already discussed then sorry - I'm not going
    > to start new discussion).
    >
    > Vadim
    >