On 4/24/05, Tom Lane wrote:
> What you are talking about is not a "read only" table, it is a
> "non-MVCC" table. This is a much greater assault on the fundamental
> semantics of Postgres than it's being painted to be in this thread.
> In particular, how is such a table going to come into being?
You need an operation that rewrites the entire table and all indexes
at the same time. That pretty much means the only sensible way would
be a special form of CLUSTER (including al the locking uglyness).
If you need to make the table READ-WRITE again you need to run CLUSTER
against it so the tuples get visibility information again and the fill
factor of the indexes is reset to 0.7.
But I have to agree with Ron that this probably makes much more sense
when you have partitioned tables and do this on partitions instead of
a full table. Archived might be a better word for it then read-only.
In response to
pgsql-hackers by date
|Next:||From: Alvaro Herrera||Date: 2005-04-25 12:41:03|
|Subject: Re: How to make lazy VACUUM of one table run in several transactions ?|
|Previous:||From: Hannu Krosing||Date: 2005-04-25 10:55:01|
|Subject: How to make lazy VACUUM of one table run in several transactions ?|