Greetings,
I would like to revisit the discussion and concur with John's perspective that
incremental progress through small, successive modifications is the appropriate
approach to move forward. Therefore, I would like to propose two distinct ideas
aimed at enhancing the functionality of inser
Hi hackers,
That's me still talking to myself :)
> Thoughts?
Evidently this works differently from what I initially thought on
Windows due to lack of fork() on this system.
PFA the patch v3. Your feedback is most welcomed.
--
Best regards,
Aleksander Alekseev
v3-0001-Clarify-the-38.10.10.-Sh
On Mon, May 22, 2023 at 6:06 PM Masahiko Sawada wrote:
>
> Thank you for updating the patch! Here are review comments:
>
> + /*
> +* Make sure that the copy command runs as the table owner, unless
> +* the user has opted out of that behaviour.
> +*/
> + run_as_o
On Sat, 20 May 2023 at 00:01, Jacob Champion wrote:
> - Some clients in the wild (psycopg2/psycopg) suppress all notifications
> during PQconnectPoll().
If there is anything we can improve in psycopg please reach out.
-- Daniele
On Tue, May 23, 2023 at 2:38 PM Richard Guo wrote:
> I came across $subject on master and here is the query I'm running.
>
> create table t (a int unique, b int);
>
> explain (costs off)
> select 1 from t t1
> left join t t2 on true
>inner join t t3 on true
> left join t t4 on t2.a =
On Mon, May 22, 2023 at 10:36 PM Masahiko Sawada wrote:
>
> On Wed, May 17, 2023 at 10:10 AM Ajin Cherian wrote:
> >
> > On Mon, May 15, 2023 at 10:47 PM Masahiko Sawada
> > wrote:
> > >
> > > On Mon, May 15, 2023 at 5:44 PM Ajin Cherian wrote:
> > > >
> > > > On Fri, May 12, 2023 at 9:55 PM A
Hi,
> [...]
> A-ha, it's because I didn't add the new test_resowner directory to the
> src/test/modules/meson.build file. Fixed.
>
> Thanks for the review!
You probably already noticed, but for the record: cfbot seems to be
extremely unhappy with the patch.
--
Best regards,
Aleksander Alekseev
Re: Andres Freund
> Move snowball_create.sql creation into perl file
>
> This is in preparation for building postgres with meson / ninja.
>
> We already have duplicated code for this between the make and msvc
> builds. Adding a third copy seems like a bad plan, thus move the generation
> into a p
Re: To Andres Freund
> this seems to have broken out-of-tree builds from tarballs:
>
> /usr/bin/install -c -m 644 snowball_create.sql
> '/srv/projects/postgresql/debian/16/build/tmp_install/usr/share/postgresql/16'
> /usr/bin/install: cannot stat 'snowball_create.sql': No such file or directory
Re: Andres Freund
> A prototype of that approach is attached. I pushed the retry handling into the
> pg_* routines where applicable. I guess we could add pg_* routines for
> FileFallocate(), FilePrewarm() etc as well, but I didn't do that here.
>
> Christoph, could you verify this fixes your issu
Christoph Berg writes:
>> this seems to have broken out-of-tree builds from tarballs:
>>
>> /usr/bin/install -c -m 644 snowball_create.sql
>> '/srv/projects/postgresql/debian/16/build/tmp_install/usr/share/postgresql/16'
>> /usr/bin/install: cannot stat 'snowball_create.sql': No such file or
>>
Re: Tom Lane
> I think the attached will do for a proper fix. I'm not inclined
> to re-wrap just for this.
Sure, I just posted it here in case others run into the same problem.
Thanks!
Christoph
Hi,
> Yeah - I think we had better reserve the fourth bit pattern for
> something extensible e.g. another byte or several to specify the
> actual method, so that we don't have a hard limit of 4 methods.
TWIMC there is an ongoing discussion [1] of making TOAST pointers
extendable since this is a d
When creating a private service for another instance of PostgreSQL I used the
template of postgresql-15.service file installed into /usr/lib/systemd/system
on Fedora 38 provided through the installation for postgres 15.3 from PGDG
repositories.
There I noticed that the line ExecStart still use
On Sat, Apr 8, 2023 at 5:26 AM Andres Freund wrote:
> TAP test for logical decoding on standby
Small nitpicks:
1. The test names generated by check_slots_conflicting_status() start
with a capital letter, while most other test names start with a
lower-case letter.
2. The function is called 7 tim
On Tue, May 23, 2023 at 4:22 AM Daniele Varrazzo
wrote:
> On Sat, 20 May 2023 at 00:01, Jacob Champion wrote:
> > - Some clients in the wild (psycopg2/psycopg) suppress all notifications
> > during PQconnectPoll().
>
> If there is anything we can improve in psycopg please reach out.
Will do, tha
Hi,
it seems there's a fairly annoying memory leak in trigger code,
introduced by
commit fc22b6623b6b3bab3cb057ccd282c2bfad1a0b30
Author: Peter Eisentraut
Date: Sat Mar 30 08:13:09 2019 +0100
Generated columns
...
which added GetAllUpdatedColumns() and uses it many places inste
On Thu, May 18, 2023 at 8:06 AM Matthias van de Meent
wrote:
> This enum still has many options to go before it exceeds the maximum
> of the uint8 va_tag field. Therefore, I don't think we have no disk
> representations left, nor do I think we'll need to add another option
> to the ToastCompressio
Tomas Vondra writes:
> it seems there's a fairly annoying memory leak in trigger code,
> introduced by
> ...
> Attached is a patch, restoring the pre-12 behavior for me.
> While looking for other places allocating stuff in ExecutorState (for
> the UPDATE case) and leaving it there, I found two mo
Hi!
I noticed that errors due to writable CTEs in read-only or non-volatile context
say the offensive command is SELECT.
For example a writeable CTE in a IMMUTABLE function:
CREATE TABLE t (x INTEGER);
CREATE FUNCTION immutable_func()
RETURNS INTEGER
LANGUAGE SQL
IMMUTABLE
AS $$
WI
Hi,
On 2023-05-23 18:23:00 +0200, Tomas Vondra wrote:
> This means that for an UPDATE with triggers, we may end up calling this
> for each row, possibly multiple bitmaps. And those bitmaps are allocated
> in ExecutorState, so won't be freed until the end of the query :-(
Ugh.
I've wondered abou
Andres Freund writes:
> I've wondered about some form of instrumentation to detect such issues
> before.
Yeah.
> Could it help to have a mode where the executor shutdown hook checks how much
> memory is allocated in ExecutorState and warns if its too much?
It'd be very hard to set a limit for w
Hello,
We (Neon) have noticed that pgbench can be quite slow to populate data
in regard to higher latency connections. Higher scale factors exacerbate
this problem. Some employees work on a different continent than the
databases they might be benchmarking. By moving pgbench to use COPY for
populat
Markus Winand writes:
> I noticed that errors due to writable CTEs in read-only or non-volatile
> context say the offensive command is SELECT.
Good point.
> My first thought was that these error messages should mention INSERT, but
> after looking into the source I’m not sure anymore. The name
Hi,
On 2023-05-23 13:28:30 -0400, Tom Lane wrote:
> Andres Freund writes:
> > Could it help to have a mode where the executor shutdown hook checks how
> > much
> > memory is allocated in ExecutorState and warns if its too much?
>
> It'd be very hard to set a limit for what's "too much", since th
Andres Freund writes:
> On 2023-05-23 13:28:30 -0400, Tom Lane wrote:
>> Why? Unlike Lists, those things are already a single palloc chunk.
> We do a fair amount of 8 byte allocations - they have quite a bit of overhead,
> even after c6e0fe1f2a0. Not needing allocations for the common case of
>
Hi,
On 2023-05-23 10:46:30 -0400, Tom Lane wrote:
> Christoph Berg writes:
> >> this seems to have broken out-of-tree builds from tarballs:
> >>
> >> /usr/bin/install -c -m 644 snowball_create.sql
> >> '/srv/projects/postgresql/debian/16/build/tmp_install/usr/share/postgresql/16'
> >> /usr/bin/
Richard Guo writes:
>> create table t (a int unique, b int);
>>
>> explain (costs off)
>> select 1 from t t1
>> left join t t2 on true
>> inner join t t3 on true
>> left join t t4 on t2.a = t4.a and t2.a = t3.a;
>> ERROR: no relation entry for relid 6
Ugh.
> Considering that clone clauses shou
[ dropping -packagers ]
Andres Freund writes:
> I guess I need to go and check how long the "release" tarball generation
> takes...
It's quick except for the documentation-generating steps. Maybe
we could test that part only once?
regards, tom lane
Hi all,
As touched on in past threads, our SCRAM implementation is slightly
nonstandard and doesn't always protect the entirety of the
authentication handshake:
- the username in the startup packet is not covered by the SCRAM
crypto and can be tampered with if channel binding is not in effect,
or
This is a short patch which cleans up code for unlogged LSNs. It replaces the
existing spinlock with
atomic ops. It could provide a performance benefit for future uses of
unlogged LSNS, but for now it is simply a cleaner implementation.
0001-Use-atomic-ops-for-unloggedLSNs.patch
Description: 0
On Mon, 22 May 2023 at 07:05, Jonathan S. Katz wrote:
> * Parallel execution of queries that use `FULL` and `OUTER` joins
I think this should be `RIGHT` joins rather than `OUTER` joins.
LEFT joins have been parallelizable I think for a long time now.
David
On Tue, 23 May 2023 at 06:04, Bruce Momjian wrote:
>
> On Mon, May 22, 2023 at 10:59:36AM -0700, Andres Freund wrote:
> > And here it's not just performance, but also memory usage, including steady
> > state memory usage.
>
> Understood. I continue to need help determining which items to include.
On 5/23/23 18:39, Tom Lane wrote:
> Tomas Vondra writes:
>> it seems there's a fairly annoying memory leak in trigger code,
>> introduced by
>> ...
>> Attached is a patch, restoring the pre-12 behavior for me.
>
>> While looking for other places allocating stuff in ExecutorState (for
>> the UP
Greetings,
* Jacob Champion (jchamp...@timescale.com) wrote:
> As touched on in past threads, our SCRAM implementation is slightly
> nonstandard and doesn't always protect the entirety of the
> authentication handshake:
>
> - the username in the startup packet is not covered by the SCRAM
> crypto
On 5/23/23 19:14, Andres Freund wrote:
> Hi,
>
> On 2023-05-23 18:23:00 +0200, Tomas Vondra wrote:
>> This means that for an UPDATE with triggers, we may end up calling this
>> for each row, possibly multiple bitmaps. And those bitmaps are allocated
>> in ExecutorState, so won't be freed until
Tomas Vondra writes:
> The really hard thing was determining what causes the memory leak - the
> simple instrumentation doesn't help with that at all. It tells you there
> might be a leak, but you don't know where did the allocations came from.
> What I ended up doing is a simple gdb script that
On Tue, May 23, 2023 at 08:24:45PM +, John Morris wrote:
> This is a short patch which cleans up code for unlogged LSNs. It
> replaces the existing spinlock with atomic ops. It could provide a
> performance benefit for future uses of unlogged LSNS, but for now
> it is simply a cleaner impleme
On 5/23/23 4:37 PM, David Rowley wrote:
On Mon, 22 May 2023 at 07:05, Jonathan S. Katz wrote:
* Parallel execution of queries that use `FULL` and `OUTER` joins
I think this should be `RIGHT` joins rather than `OUTER` joins.
LEFT joins have been parallelizable I think for a long time now.
I
On Tue, May 23, 2023 at 05:56:13PM +0300, Aleksander Alekseev wrote:
> TWIMC there is an ongoing discussion [1] of making TOAST pointers
> extendable since this is a dependency for several patches that are
> currently in development.
Thanks for the ping. I have seen and read the other thread, and
On 5/22/23 4:18 PM, Robert Haas wrote:
On Sun, May 21, 2023 at 3:05 PM Jonathan S. Katz wrote:
* Support for regular expressions for matching usernames and databases
names in `pg_hba.conf`, and user names in `pg_ident.conf`
I suggest that this is not a major feature.
Perhaps the work that I
I wrote:
> the current insert/delete paths are quite complex. Using bitmap heap scan
as a motivating use case, I hope to refocus complexity to where it's most
needed, and aggressively simplify where possible.
Sometime in the not-too-distant future, I will start a new thread focusing
on bitmap heap
On Tue, May 23, 2023 at 12:33:50PM -0400, Robert Haas wrote:
> For projects like this, the details matter a lot. If the goal is to
> add a new compression type that behaves like the existing compression
> types, more or less, then I think we should allocate the last
> ToastCompressionId bit to mean
On Mon, May 22, 2023 at 10:02:02AM -0500, Tristan Partin wrote:
> The way that pgbench handled SIGINT changed in
> 1d468b9ad81b9139b4a0b16b416c3597925af4b0. Unfortunately this had a
> couple of unintended consequences, at least from what I can tell[1].
>
> - CTRL-C no longer stops the program unle
Thanks all for the feedback. It was a nice idea and it *almost*
works, but it seems like we just can't drop segmented mode. And the
automatic transition schemes I showed don't make much sense without
that goal.
What I'm hearing is that something simple like this might be more acceptable:
* init
On Mon, May 22, 2023 at 12:26:34PM -0500, Tristan Partin wrote:
> This is a way that would solve bug #17698[1]. It just reuses the same
> handler as SIGINT (with a function rename).
>
> This patch works best if it is combined with my previous submission[2].
> I can rebase that submission if and wh
On Mon, May 22, 2023 at 06:48:37PM +0900, torikoshia wrote:
> While working on [1], I thought there seems to be unnecessary #include
> "pg_getopt.h".
> getopt_long.h has already included pg_getopt.h, but some files include both
> getopt.h and getopt_long.h.
Right, these could be removed. I am not
On Mon, May 22, 2023 at 06:24:49PM +0900, torikoshia wrote:
> Thanks for your advice, attached patches.
0001 looks OK, thanks!
+Remove files including backup history file.
This could be reworded as "Remove backup history files.", I assume.
+ Note that when oldestkeptwalfile is a
ba
On Tue, May 23, 2023 at 05:02:50PM -0400, Stephen Frost wrote:
> * Jacob Champion (jchamp...@timescale.com) wrote:
>> As touched on in past threads, our SCRAM implementation is slightly
>> nonstandard and doesn't always protect the entirety of the
>> authentication handshake:
>>
>> - the username i
Hi,
On 2023-05-24 09:59:18 +0900, Michael Paquier wrote:
> On Mon, May 22, 2023 at 06:48:37PM +0900, torikoshia wrote:
> > While working on [1], I thought there seems to be unnecessary #include
> > "pg_getopt.h".
> > getopt_long.h has already included pg_getopt.h, but some files include both
> > g
On Tue, May 23, 2023 at 04:25:59PM +0200, Christoph Berg wrote:
> I believe this issue is still open for PG16.
Right. I've added an item to the list, to not forget.
--
Michael
signature.asc
Description: PGP signature
Greetings,
* Michael Paquier (mich...@paquier.xyz) wrote:
> On Tue, May 23, 2023 at 05:02:50PM -0400, Stephen Frost wrote:
> > * Jacob Champion (jchamp...@timescale.com) wrote:
> >> As touched on in past threads, our SCRAM implementation is slightly
> >> nonstandard and doesn't always protect the
On Tue, May 23, 2023 at 09:46:58PM -0400, Stephen Frost wrote:
> Not without breaking things we support today and for what seems like an
> unclear benefit given that we've got channel binding today (though
> perhaps we need to make sure there's ways to force it on both sides to
> be on and to encou
At Wed, 26 Apr 2023 11:37:55 +1200, Thomas Munro wrote
in
> On Tue, Apr 25, 2023 at 12:16 PM Andres Freund wrote:
> > On 2023-04-24 15:32:25 -0700, Andres Freund wrote:
> > > We obviously can add a retry loop to FileFallocate(), similar to what's
> > > already present e.g. in FileRead(). But I
Greetings,
* Michael Paquier (mich...@paquier.xyz) wrote:
> On Tue, May 23, 2023 at 09:46:58PM -0400, Stephen Frost wrote:
> > Not without breaking things we support today and for what seems like an
> > unclear benefit given that we've got channel binding today (though
> > perhaps we need to make
On Tue, May 23, 2023 at 4:10 PM Benjamin Coutu wrote:
>
> Greetings,
>
> I would like to revisit the discussion and concur with John's perspective
that incremental progress through small, successive modifications is the
appropriate approach to move forward. Therefore, I would like to propose
two d
Hi,
On 2023-05-24 10:56:28 +0900, Kyotaro Horiguchi wrote:
> At Wed, 26 Apr 2023 11:37:55 +1200, Thomas Munro
> wrote in
> > On Tue, Apr 25, 2023 at 12:16 PM Andres Freund wrote:
> > > On 2023-04-24 15:32:25 -0700, Andres Freund wrote:
> > > > We obviously can add a retry loop to FileFallocate
On Tue, May 23, 2023 at 07:50:36AM +0900, Michael Paquier wrote:
> Seeing the precedent with --no-blobs and --blobs, yes, that should be
> enough. You may want to wait until beta1 is stamped to apply
> something, though, as the period between the stamp and the tag is used
> to check the state of t
Hi, and thanks for the patch! It is an interesting idea.
I have not yet fully read this thread, so below are only my first
impressions after looking at patch 0001. Sorry if some of these were
already discussed earlier.
TBH the patch "reuse-workers" logic seemed more complicated than I had
imagine
On Wed, May 24, 2023 at 08:37:45AM +1200, David Rowley wrote:
> On Mon, 22 May 2023 at 07:05, Jonathan S. Katz wrote:
> > * Parallel execution of queries that use `FULL` and `OUTER` joins
>
> I think this should be `RIGHT` joins rather than `OUTER` joins.
>
> LEFT joins have been parallelizable
On Tue, May 23, 2023 at 06:27:23PM -0400, Jonathan Katz wrote:
> On 5/23/23 4:37 PM, David Rowley wrote:
> > On Mon, 22 May 2023 at 07:05, Jonathan S. Katz wrote:
> > > * Parallel execution of queries that use `FULL` and `OUTER` joins
> >
> > I think this should be `RIGHT` joins rather than `OUTE
On Tue, May 23, 2023 at 12:14:04PM +0700, John Naylor wrote:
> On Tue, May 23, 2023 at 11:26 AM Bruce Momjian wrote:
> > > > Allow xid/subxid searches to use vector operations on x86-64
> architectures
> > > (Nathan Bossart)
> > >
> > > When moved to the performance section, it would be something
On Wed, 24 May 2023 at 15:54, Bruce Momjian wrote:
>
> On Wed, May 24, 2023 at 08:37:45AM +1200, David Rowley wrote:
> > On Mon, 22 May 2023 at 07:05, Jonathan S. Katz wrote:
> > > * Parallel execution of queries that use `FULL` and `OUTER` joins
> >
> > I think this should be `RIGHT` joins rathe
At Tue, 23 May 2023 19:28:45 -0700, Andres Freund wrote in
> Hi,
>
> On 2023-05-24 10:56:28 +0900, Kyotaro Horiguchi wrote:
> > At Wed, 26 Apr 2023 11:37:55 +1200, Thomas Munro
> > wrote in
> > > On Tue, Apr 25, 2023 at 12:16 PM Andres Freund wrote:
> > > > On 2023-04-24 15:32:25 -0700, Andr
On Tue, May 23, 2023 at 12:14:04PM +0700, John Naylor wrote:
> On Tue, May 23, 2023 at 11:26 AM Bruce Momjian wrote:
> >
> > On Tue, May 23, 2023 at 09:58:30AM +0700, John Naylor wrote:
> > > > Allow ASCII string detection to use vector operations on x86-64
> architectures
> > > (John Naylor)
> >
On Tue, May 23, 2023 at 10:01:03PM -0400, Stephen Frost wrote:
> To the extent that there was an issue when it was implemented ... it's
> now been implemented and so that was presumably overcome (though I don't
> really specifically recall what the issues were there? Seems like it
> wouldn't matte
On Wed, May 24, 2023 at 11:19 AM Bruce Momjian wrote:
>
> Second, you might be correct that the section is wrong. I thought of
> CPU instructions as something tied to the compiler, so part of the build
> process or source code, but the point we should be make is that we have
> these acceleration,
On Mon, 22 May 2023 at 11:27, shveta malik wrote:
>
> On Wed, May 17, 2023 at 4:45 PM vignesh C wrote:
> >
> > On Wed, 17 May 2023 at 15:41, shveta malik wrote:
> > >
> > > On Fri, May 12, 2023 at 12:03 PM shveta malik
> > > wrote:
> > > >
> > > > On Tue, May 9, 2023 at 4:23 PM shveta malik
On Tuesday, May 23, 2023 1:53 AM Andres Freund wrote:
> On 2023-05-22 12:15:07 +, Zhijie Hou (Fujitsu) wrote:
> > About "a backend doing logical decoding", do you mean the case when a
> user
> > start a backend and invoke pg_logical_slot_get_changes() to do the logical
> > decoding ? If so, it
On Wed, May 24, 2023 at 2:48 AM Tom Lane wrote:
> Richard Guo writes:
> > Considering that clone clauses should always be outer-join clauses not
> > filter clauses, I'm wondering if we can add an additional check for that
> > in RINFO_IS_PUSHED_DOWN, something like
>
> > #define RINFO_IS_PUSHED
> That's worth trying out. It might also then be worth trying to push both
> unordered values -- the big one up / the small one down. I've seen other
> implementations do that, but don't remember where, or what it's called.
It is important that we do not do 2 compares two avoid one copy (assignm
On 24.05.23 02:34, Thomas Munro wrote:
Thanks all for the feedback. It was a nice idea and it *almost*
works, but it seems like we just can't drop segmented mode. And the
automatic transition schemes I showed don't make much sense without
that goal.
What I'm hearing is that something simple li
72 matches
Mail list logo
