Hi all,
this is a little controversial, but hagin a function defined as
immutable that selects a record out of a table, it is impossible to
create an index over such function if the column has a constraint like
a primary key.
Here it is the use case:
DROP TABLE if exists t;
drop table if exist
Hi Ron,
> On 11 Mar 2025, at 20:34, Ron Johnson wrote:
>
> If you don't do much DDL, the LR should be quite workable.
DDL during logical replication unfortunately is a show-stopper.
Cheers,
Paul
Luca Ferrari writes:
> On Wed, Mar 12, 2025 at 12:54 PM Artur Zakirov wrote:
>> In your case `base/357283/365810` file is a new index file. For some
>> reason Postgres tries to read the new index. I suppose this is because
>> during reading the table `t` within the function `f_t` it tries to
>> a
Hey,
On Wed, 12 Mar 2025 at 10:11, Luca Ferrari wrote:
> Now, according to the documentation, the function f_t is immutable
> since it is not modifying the database, so what is going on? And why
> is the same function working if the table has not the constraint on
> the column?
I can reproduce t
On 3/12/25 14:31, Luca Ferrari wrote:
On Wed, Mar 12, 2025 at 12:54 PM Artur Zakirov wrote:
I can reproduce this with the table `t` on PG 15.10.
I didn't mention I'm running 16.6, but I'm pretty sure it is
reproducible on other versions too.
In your case `base/357283/365810` file is a new in
On 3/12/25 08:46, Adrian Klaver wrote:
On 3/11/25 13:24, Adrian Klaver wrote:
On 3/11/25 12:55, mark bradley wrote:
It happened again. Now there are no sequences (although there once
was).
Read my previous post and provide the information requested.
Mark sent me the below, which answers s
Hi Greg,
> On 12 Mar 2025, at 21:31, Greg Sabino Mullane wrote:
>
> Keep in mind that you only need to reindex text-based indexes. Ints are still
> gonna int. So it might not be too bad.
Yes, I know, but unfortunately the worst case index of them all will still take
a few hours.
Cheers,
Paul
On Wed, Mar 12, 2025 at 4:16 PM Paul Foerster
wrote:
> Hi Ron,
>
> > On 12 Mar 2025, at 17:59, Ron Johnson wrote:
> >
> > Developers making DDL changes on production databases?
>
> Of course not. But I can't block developer databases. That'd make a few
> hundred developers happy.
>
> > Or are th
On 3/12/25 16:55, Paul Foerster wrote:
Hi Adrian,
On 12 Mar 2025, at 21:26, Adrian Klaver wrote:
A good summary of the glibc issue:
https://wiki.postgresql.org/wiki/Locale_data_changes
With distro information:
https://wiki.postgresql.org/wiki/Locale_data_changes#What_Linux_distributions_ar
On 3/11/25 13:24, Adrian Klaver wrote:
On 3/11/25 12:55, mark bradley wrote:
It happened again. Now there are no sequences (although there once was).
Read my previous post and provide the information requested.
Mark sent me the below, which answers some of the questions, namely
there is i
On Wed, Mar 12, 2025 at 12:54 PM Artur Zakirov wrote:
>
> I can reproduce this with the table `t` on PG 15.10.
I didn't mention I'm running 16.6, but I'm pretty sure it is
reproducible on other versions too.
>
> In your case `base/357283/365810` file is a new index file. For some
> reason Postgr
Hi Christophe,
> On 12 Mar 2025, at 12:16, Christophe Pettus wrote:
>
> You *can* apply DDL while logical replication is going on, as long as you do
> so in a disciplined way. This generally means applying it to the subscriber
> before you apply it to the publisher, and making sure that any c
On Wed, Mar 12, 2025 at 12:48 PM Paul Foerster
wrote:
> Hi Christophe,
>
> > On 12 Mar 2025, at 12:16, Christophe Pettus wrote:
> >
> > You *can* apply DDL while logical replication is going on, as long as
> you do so in a disciplined way. This generally means applying it to the
> subscriber be
On Tue, Mar 11, 2025 at 2:35 PM Paul Foerster
wrote:
> The question was a bit of an idea. So the glibc version in not known yet,
> but I'm highly confident that they will differ. A reindex could in theory
> be possible in most cases, but is a definite show stopper on some of our
> databases, beca
Hi Adrian,
> On 12 Mar 2025, at 21:26, Adrian Klaver wrote:
>
> A good summary of the glibc issue:
>
> https://wiki.postgresql.org/wiki/Locale_data_changes
>
> With distro information:
>
> https://wiki.postgresql.org/wiki/Locale_data_changes#What_Linux_distributions_are_affected
I know the a
Hi Ron,
> On 12 Mar 2025, at 21:50, Ron Johnson wrote:
>
> No, I think it's 100% on point: logically replicate the Prod databases, while
> pg_dump/pg_restore of the dev and pre-prod databases happen on weekends.
Yes, I live for and in the company. 🤣 SNCR.
No, seriously, I'm one of only two Po
On 3/11/25 11:34, Paul Foerster wrote:
Hi Devrim, Thomas, Adrian, Ron, Joe,
answering to myself as answering to five postings in one go is impossible. 🤣
Are there any obstacles that definitely make that a no-go? Do I risk
corruption? It's both Linux, just a different distribution.
The quest
On 3/12/25 13:55, Paul Foerster wrote:
Hi Adrian,
On 12 Mar 2025, at 21:26, Adrian Klaver wrote:
A good summary of the glibc issue:
https://wiki.postgresql.org/wiki/Locale_data_changes
With distro information:
https://wiki.postgresql.org/wiki/Locale_data_changes#What_Linux_distributions_ar
18 matches
Mail list logo
