Re: [HACKERS] Hot Backup Ability - Mailing list pgsql-hackers

From Michael Richards
Subject Re: [HACKERS] Hot Backup Ability
Date
Msg-id Pine.BSF.4.10.9906292146050.12242-100000@scifair.acadiau.ca
Whole thread Raw
In response to Re: [HACKERS] Hot Backup Ability  (Duane Currie <dcurrie@sandman.acadiau.ca>)
Responses Re: [HACKERS] Hot Backup Ability
List pgsql-hackers
On Tue, 29 Jun 1999, Duane Currie wrote:

> Anyway, version 6.5 apparently supports hot backups by using MVCC
> to give a view of a consistent version of the database during a
> pg_dump (http://postgresql.nextpath.com/doxlist.html  search for
'backup')
Hrm. Nothing pops out.   
Just out of curiosity, I did a DUMP on the database while running a script
that ran a pile of updates. When I restored the database files, it was so
corrupted that I couldn't even run a select. vacuum just core dumped...   
If I can just run pg_dump to back it up, how does this guarantee any sort
of referential integrity? Also during such a dump, it seems that things   
block while waiting for a lock. This also happens during a pg_vacuum. I   
thought that mvcc was supposed to stop this...   
-Michael   



pgsql-hackers by date:

Previous
From: "Hiroshi Inoue"
Date:
Subject: RE: [HACKERS] 6.5.1 status
Next
From: Bruce Momjian
Date:
Subject: Re: [HACKERS] Hot Backup Ability