Hello,
Forgive me if this is a dopey question. I'm running a web app with a
dB that takes me a half hour to regenerate. Instead of closing down
every day, I'd like insert the data into a temp table "stuff_tmp" and
then rename that to the permanent table "stuff", so the application is
not off-line for so long.
The table "stuff_tmp" has a primary key constraint. When I rename the
table, this constraint does not get renamed-- it continues to have the
name "stuff_tmp_pkey"-- and (you guessed it) the next time I run the
script pg complains that it can't make "stuff_tmp" because the
constraint already exists.
That I can see, I can't rename the constraint. Do I have that correct?
So I thought to drop the constraint. That I can see I can't add a
primary key constraint "stuff_pkey". Is that correct?
Can I simulate (sort of) a primary key constraint by adding a UNIQUE
index, and a NOT NULL check? That is, if I add those two, do I lose
anything compared with the original primary key constraint?
Thank you,
Jim