On 30 November 2015 at 17:01, Bruce Momjian <bruce@xxxxxxxxxx> wrote:
On Mon, Nov 30, 2015 at 04:51:15PM +0000, Benedikt Grundmann wrote:
> Are you able to compile from 9.4 git head and test that? It seems
> dumping inheriting constraints from parents has not worked properly for
> some time.
>
>
> Do I need to get the latest/head 9.2 or the latest/head 9.4 or both? For what
> it is worth I just tried after upgrading to the latest released 9.2 (and same
> 9.45) and that didn't work :-(
You actually need non-released 9.4.X code that is in pg_dump, and we use
9.4 pg_dump to dump the 9.2 database.
> I should certainly be able to compile from source. But the upgrade to 9.4 is
> by far not high on my priority stack (other than maybe some speed wins there is
> nothing in 9.4 that we are eager for, there are some niceties but I can happily
> live without all of them for years) and has already consumed way more time than
> I had scheduled for it. So I'll return to focus on other work for at least
> this week and maybe more depending on how that work goes.
>
> Thanks to everyone I'll certainly update this thread if / when I have more time
> to devote to this.
The simplest solution is to wait for 9.4.6 to be released and test that.
I just tried this again. This time from 9.2.17 to 9.5.4 and pg_upgrade chokes with this:
[root@igm-dbc-001 upgrade-logs]# tail pg_upgrade_dump_16416.log
pg_restore: [archiver (db)] could not execute query: ERROR: syntax error at or near "=>"
LINE 1: CREATE OPERATOR => (
^
Command was: CREATE OPERATOR => (
PROCEDURE = "tconvert",
LEFTARG = "text",
RIGHTARG = "text"
);
-- For binary upgrade, hand...
Any pointers are appreciated.
Thanks,
Bene
--
Bruce Momjian <bruce@xxxxxxxxxx> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ As you are, so once was I. As I am, so you will be. +
+ Roman grave inscription +
