Sequence Start number not dumped correctly

Started by Reto Stammover 23 years ago3 messagesbugs
Jump to latest
#1Reto Stamm
reto@retostamm.com

Hi,

When I have a sequence like this:

CREATE SEQUENCE "test_id_seq" start -32768 increment 1 maxvalue 32767 minvalue
-32768 cache 1;

and dump it using

pg_dump -s -c -f schema.txt testdb

it will show up in schema.txt like this:

CREATE SEQUENCE "test_id_seq" start 1 increment 1 maxvalue 32767 minvalue -32768
cache 1;

This makes me think that the start value is hardwired somewhere, and I wanted to
let you know. PostgreSQL is great! I use it for everything!

Thanks a lot

--reto

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: Reto Stamm (#1)
Re: Sequence Start number not dumped correctly

Reto Stamm <reto@retostamm.com> writes:

When I have a sequence like this:
CREATE SEQUENCE "test_id_seq" start -32768 increment 1 maxvalue 32767 minvalue
-32768 cache 1;

it will show up in schema.txt like this:

CREATE SEQUENCE "test_id_seq" start 1 increment 1 maxvalue 32767 minvalue -32768
cache 1;

How old is your Postgres? It works here.

But note that there is probably also a select setval() for the sequence,
which is what *really* determines the starting value.

regards, tom lane

#3Reto Stamm
reto@retostamm.com
In reply to: Tom Lane (#2)
Re: Sequence Start number not dumped correctly

Hello Tom,

I am sorry, I can not reproduce it any more either. I must have made a mistake.

It works perfectly (like the rest of Postgres). Thank you very much, I am sorry
for the waste of time.

Best regards

--reto

Quoting Tom Lane <tgl@sss.pgh.pa.us>:

Show quoted text

Reto Stamm <reto@retostamm.com> writes:

When I have a sequence like this:
CREATE SEQUENCE "test_id_seq" start -32768 increment 1 maxvalue 32767

minvalue

-32768 cache 1;

it will show up in schema.txt like this:

CREATE SEQUENCE "test_id_seq" start 1 increment 1 maxvalue 32767 minvalue

-32768

cache 1;

How old is your Postgres? It works here.

But note that there is probably also a select setval() for the sequence,
which is what *really* determines the starting value.

regards, tom lane