Hi,
On Fri, Jan 05, 2024 at 12:11:44AM -0600, Nathan Bossart wrote:
> On Wed, Jan 03, 2024 at 07:59:45AM +, Bertrand Drouvot wrote:
> > +1 to add a test and put in a place that would produce failures at build
> > time.
> > I think that having the test in the script that generates the header
On Fri, Jan 5, 2024 at 12:49 PM Michael Paquier wrote:
>
> I mean why? We test a bunch of stuff in src/test/modules/, and this
> is not intended to be released to the outside world.
>
> Putting that in contrib/ has a lot of extra cost. One is
> documentation and more complexity regarding
On Fri, Jan 5, 2024 at 6:26 AM Jeff Davis wrote:
>
> > 2. Can one use {FDW, user_mapping, foreign_server} combo other than
> > the built-in pg_connection_fdw?
>
> Yes, you can use any FDW for which you have USAGE privileges, passes
> the validations, and provides enough of the expected fields to
On Fri, Jan 05, 2024 at 12:41:33PM +0530, Ashutosh Bapat wrote:
> Well, you have already showed that the SQL interface created for the
> test module is being used for testing a core feature. The tests for
> that should stay somewhere near the other tests for those features.
> Using an extension
On Fri, Jan 5, 2024 at 5:08 AM Michael Paquier wrote:
>
> >> I suggest we move test_injection_points from src/test/modules to
> >> contrib/ and rename it as "injection_points". The test files may still
> >> be named as test_injection_point. The TAP tests in 0003 and 0004 once
> >> moved to their
On Thu, 4 Jan 2024 at 16:46, Amit Kapila wrote:
>
> On Thu, Jan 4, 2024 at 12:22 PM Shlok Kyal wrote:
> >
> > Hi,
> > I was testing the patch with following test cases:
> >
> > Test 1 :
> > - Create a 'primary' node
> > - Setup physical replica using pg_basebackup "./pg_basebackup –h
> >
On Wed, Jan 03, 2024 at 11:34:25AM +0900, Michael Paquier wrote:
> On Tue, Jan 02, 2024 at 11:31:20AM -0600, Nathan Bossart wrote:
>> +# Find the location of lwlocknames.h.
>> +my $include_dir = $node->config_data('--includedir');
>> +my $lwlocknames_file =
On Wed, Jan 03, 2024 at 07:59:45AM +, Bertrand Drouvot wrote:
> +1 to add a test and put in a place that would produce failures at build time.
> I think that having the test in the script that generates the header file is
> more
> appropriate (as building the documentation looks less usual to
> On 3 Jan 2024, at 18:22, Jelte Fennema-Nio wrote:
>
>
>> In my case I have scripts that I want to execute with limited privileges
>> and make sure the scripts cannot escape the sandbox via RESET ROLE.
>
> Depending on the desired workflow I think that could work for you too.
> Because it
On Fri, Jan 5, 2024 at 2:38 PM Hayato Kuroda (Fujitsu)
wrote:
...
> 2.
> I'm not sure it should be listed as step 10. I felt that it should be new
> section.
> At that time other steps like "Prepare for {publisher|subscriber} upgrades"
> can be moved as well.
> Thought?
During my review, I
Here are some review comments for patch v1-0001.
==
doc/src/sgml/ref/pgupgrade.sgml
1. GENERAL - blank lines
Most (but not all) of your procedure steps are preceded by blank lines
to make them more readable in the SGML. Add the missing blank lines
for the steps that didn't have them.
2.
On Tue, Jan 2, 2024 at 9:59 AM Paul Jungwirth
wrote:
>
> On 12/31/23 00:51, Paul Jungwirth wrote:
> > That's it for now.
>
> Here is another update. I fixed FOR PORTION OF on partitioned tables, in
> particular when the attnums
> are different from the root partition.
>
> Rebased to cea89c93a1.
On Thu, Jan 4, 2024 at 9:27 PM Euler Taveira wrote:
>
> On Thu, Jan 4, 2024, at 3:05 AM, Amit Kapila wrote:
>
> Won't it be a better user experience that after setting up the target
> server as a logical replica (subscriber), it started to work
> seamlessly without user intervention?
>
>
> If we
čt 4. 1. 2024 v 22:02 odesílatel Tom Lane napsal:
> Pavel Stehule writes:
> > Now, I think so this simple patch is ready for committers
>
> I pushed this with some editorialization -- mostly, rewriting the
> documentation and comments. I found that the existing docs for %TYPE
> were not great.
On Thu, Jan 4, 2024 at 9:18 PM Euler Taveira wrote:
>
> On Thu, Jan 4, 2024, at 2:41 AM, Amit Kapila wrote:
>
>
> I think asking users to manually remove such slots won't be a good
> idea. We might want to either remove them by default or provide an
> option to the user.
>
>
> Am I correct that
On Sun, 29 Oct 2023 at 11:14, Hayato Kuroda (Fujitsu)
wrote:
>
> Dear Andres,
>
> While tracking BF failures related with pg_ugprade, I found the same failure
> has still happened [1] - [4].
> According to the log, the output directory was remained even after the
> successful upgrade [5].
> I
On Fri, Jan 5, 2024 at 8:59 AM shveta malik wrote:
>
> On Thu, Jan 4, 2024 at 7:24 PM Bertrand Drouvot
> wrote:
> >
> > 4 ===
> >
> > Looking closer, the only place where walrcv_connect() is called with
> > replication
> > set to false and logical set to false is in ReplSlotSyncWorkerMain().
>
On Thu, Jan 4, 2024 at 5:30 PM Alexander Lakhin wrote:
>
> 03.01.2024 14:42, Amit Kapila wrote:
> >
>
> >> And the internal process is ... background writer (BgBufferSync()).
> >>
> >> So, I tried just adding bgwriter_lru_maxpages = 0 to postgresql.conf and
> >> got 20 x 10 tests passing.
> >>
>
Dear Vignesh,
Thanks for making a patch! Below part is my comments.
1.
Only two steps were added an id, but I think it should be for all the steps.
See [1].
2.
I'm not sure it should be listed as step 10. I felt that it should be new
section.
At that time other steps like "Prepare for
On Thu, Jan 4, 2024 at 7:24 PM Bertrand Drouvot
wrote:
>
> Hi,
>
> On Thu, Jan 04, 2024 at 10:27:31AM +0530, shveta malik wrote:
> > On Thu, Jan 4, 2024 at 9:18 AM shveta malik wrote:
> > >
> > > On Wed, Jan 3, 2024 at 6:33 PM Zhijie Hou (Fujitsu)
> > > wrote:
> > > >
> > > > On Tuesday,
Hi,
Andres Freund writes:
>
> On 2024-01-04 14:59:06 +0800, Andy Fan wrote:
>> My question is if someone doesn't obey the rule by mistake (everyone
>> can make mistake), shall we PANIC on a production environment? IMO I
>> think it can be a WARNING on a production environment and be a stuck
Hi,
On 2024-01-04 17:03:18 -0600, Jim Nasby wrote:
> On 1/4/24 10:33 AM, Tom Lane wrote:
>
> Robert Haas writes:
>
> On Thu, Jan 4, 2024 at 10:22 AM Tom Lane wrote:
>
> We should be making an effort to ban coding patterns like
> "return with spinlock still
On Wed, Jan 3, 2024 at 10:29 PM Nathan Bossart wrote:
> If the requirement is that normal builds use AVX2, then I fear we will be
> waiting a long time. IIUC the current proposals (building multiple
> binaries or adding a configuration option that maps to compiler flags)
> would still be opt-in,
On Wed, 2023-12-20 at 15:36 +0530, Bharath Rupireddy wrote:
> Thanks. Attaching remaining patches as v18 patch-set after commits
> c3a8e2a7cb16 and 766571be1659.
Comments:
I still think the right thing for this patch is to call
XLogReadFromBuffers() directly from the callers who need it, and not
Hi,
Attached a patch with a (hopefully) better wording of the comment.
I have unsuccessfully tried to find an official source for this policy.
So for reference some discussions about the topic:
-
https://stackoverflow.com/questions/11491065/linking-with-versioned-shared-library-in-android-ndk
-
On Thu, Jan 04, 2024 at 04:00:01PM +0300, Alexander Lakhin wrote:
> Reproduced here.
Did you just make the run slow enough to show the failure with
valgrind?
> As I can see in the failure logs you referenced, the first problem is:
> # Failed test 'inactiveslot slot invalidation is logged with
Paul Jungwirth writes:
> Here is a patch with an improved test. With the old "10" estimate we get a
> Merge Join, but now that
> the planner can see there are only ~4 elements per array, we get a Nested
> Loop.
Pushed with minor editorialization. I ended up not using the test
case, because I
On Thu, Jan 04, 2024 at 04:31:02PM -0600, Nathan Bossart wrote:
> On Thu, Jan 04, 2024 at 06:04:20PM +0530, Ashutosh Bapat wrote:
>> 0003 and 0004 are using the extension in this module for some serious
>> testing. The name of the extension test_injection_point indicates that
>> it's for testing
On 1/4/24 2:23 PM, Andres Freund wrote:
On 2024-01-02 12:36:18 -0500, Melanie Plageman wrote:
Subject: [PATCH v2 1/6] lazy_scan_skip remove unnecessary local var rel_pages
Subject: [PATCH v2 2/6] lazy_scan_skip remove unneeded local var
nskippable_blocks
I think these
Hi,
On 2024-01-04 14:59:06 +0800, Andy Fan wrote:
> My question is if someone doesn't obey the rule by mistake (everyone
> can make mistake), shall we PANIC on a production environment? IMO I
> think it can be a WARNING on a production environment and be a stuck
> when 'ifdef
On Thu, Jan 4, 2024 at 12:31 PM Robert Haas wrote:
>
> On Wed, Dec 27, 2023 at 11:27 AM Melanie Plageman
> wrote:
> > Do you have specific concerns about its correctness? I understand it
> > is an area where we have to be sure we are correct. But, to be fair,
> > that is true of all the pruning
On 1/4/24 10:33 AM, Tom Lane wrote:
Robert Haas writes:
On Thu, Jan 4, 2024 at 10:22 AM Tom Lane wrote:
We should be making an effort to ban coding patterns like
"return with spinlock still held", because they're just too prone
to errors similar to this one.
I agree
Hi,
On 2024-01-04 12:04:07 -0500, Robert Haas wrote:
> On Thu, Jan 4, 2024 at 11:33 AM Tom Lane wrote:
> > I believe it's (a). No matter what the reason for a stuck spinlock
> > is, the only reliable method of getting out of the problem is to
> > blow things up and start over. The patch
First of all, I'm a huge fan of UUID v7. So I'm very excited that this
is progressing. I'm definitely going to look closer at this patch
soon. Some tiny initial feedback:
(bikeshed) I'd prefer renaming `get_uuid_v7_time` to the shorter
`uuid_v7_time`, the `get_` prefix seems rarely used in
On Mon, Nov 20, 2023 at 10:39:43PM -0600, Nathan Bossart wrote:
> Alright. The next minor release isn't until February, so I'll let this one
> sit a little while longer in case anyone objects to back-patching something
> like this [0].
>
> [0]
On 1/3/24 5:57 PM, Cedric Villemain
wrote:
for 15 years pgfincore has been sitting quietly and being used
in large setups to help in HA and resources management.
It can perfectly stay as is, to be honest I was expecting to one
day
Thanks for the review!
On Thu, Jan 4, 2024 at 3:03 PM Andres Freund wrote:
>
> On 2023-11-17 18:12:08 -0500, Melanie Plageman wrote:
> > Assert(ItemIdIsNormal(lp));
> > htup = (HeapTupleHeader) PageGetItem(dp, lp);
> > @@ -715,7 +733,17 @@ heap_prune_chain(Buffer
On Thu, Jan 04, 2024 at 06:04:20PM +0530, Ashutosh Bapat wrote:
> 0003 and 0004 are using the extension in this module for some serious
> testing. The name of the extension test_injection_point indicates that
> it's for testing injection points and not for some serious use of
> injection callbacks
On Thu, Jan 04, 2024 at 08:53:11AM +0900, Michael Paquier wrote:
> On Tue, Jan 02, 2024 at 11:14:56PM -0600, Nathan Bossart wrote:
>> I'm wondering how important it is to cache the callbacks locally.
>> load_external_function() won't reload an already-loaded library, so AFAICT
>> this is
Committed after some additional light edits. Thanks for the patch!
On Wed, Jan 03, 2024 at 11:36:36PM +0100, Jelte Fennema-Nio wrote:
> On Wed, 3 Jan 2024 at 23:13, Tom Lane wrote:
>> I like Nathan's wording.
>
> To be clear, I don't want to block this patch on the wording of that
> single
On 1/3/24 5:47 PM, Nico Williams wrote:
Though maybe `NO RESET` isn't really needed to build these, since after
all one could use an unprivileged role and a SECURITY DEFINER function
that does the `SET ROLE` following some user-defined authentication
method, and so what
Pavel Stehule writes:
> Now, I think so this simple patch is ready for committers
I pushed this with some editorialization -- mostly, rewriting the
documentation and comments. I found that the existing docs for %TYPE
were not great. There are two separate use-cases, one for referencing
a table
Hi,
On 2024-01-02 12:36:18 -0500, Melanie Plageman wrote:
> Subject: [PATCH v2 1/6] lazy_scan_skip remove unnecessary local var rel_pages
> Subject: [PATCH v2 2/6] lazy_scan_skip remove unneeded local var
> nskippable_blocks
I think these may lead to worse code - the compiler has to reload
Hi,
On 2023-11-17 18:12:08 -0500, Melanie Plageman wrote:
> diff --git a/src/backend/access/heap/heapam.c
> b/src/backend/access/heap/heapam.c
> index 14de8158d49..b578c32eeb6 100644
> --- a/src/backend/access/heap/heapam.c
> +++ b/src/backend/access/heap/heapam.c
> @@ -8803,8 +8803,13 @@
Hi,
On 2024-01-04 12:31:36 -0500, Robert Haas wrote:
> On Wed, Dec 27, 2023 at 11:27 AM Melanie Plageman
> wrote:
> > Do you have specific concerns about its correctness? I understand it
> > is an area where we have to be sure we are correct. But, to be fair,
> > that is true of all the pruning
On 25.12.23 13:10, Amul Sul wrote:
> I think we can't support that (like alter type) since we need to
place
> this new
> pass before AT_PASS_OLD_INDEX & AT_PASS_OLD_CONSTR to re-add
indexes and
> constraints for the validation.
Could we have AT_PASS_ADD_COL
On Tue, Jan 02, 2024 at 10:11:23AM -0600, Nathan Bossart wrote:
> (In case it isn't clear, I'm volunteering to set up such a buildfarm
> machine.)
I set up "akepa" to run with -march=x86-64-v3.
--
Nathan Bossart
Amazon Web Services: https://aws.amazon.com
On Wed, Dec 27, 2023 at 11:27 AM Melanie Plageman
wrote:
> Do you have specific concerns about its correctness? I understand it
> is an area where we have to be sure we are correct. But, to be fair,
> that is true of all the pruning and vacuuming code.
I'm kind of concerned that 0002 might be a
On Thu, Jan 4, 2024 at 11:33 AM Tom Lane wrote:
> I believe it's (a). No matter what the reason for a stuck spinlock
> is, the only reliable method of getting out of the problem is to
> blow things up and start over. The patch proposed at the top of this
> thread would leave the system unable
Robert Haas writes:
> On Thu, Jan 4, 2024 at 10:22 AM Tom Lane wrote:
>> We should be making an effort to ban coding patterns like
>> "return with spinlock still held", because they're just too prone
>> to errors similar to this one.
> I agree that we don't want to add overhead, and also about
On Thu, Jan 4, 2024 at 10:22 AM Tom Lane wrote:
> I'm not a fan of adding overhead to such a performance-critical
> thing as spinlocks in order to catch coding errors that are easily
> detectable statically. IMV the correct usage of spinlocks is that
> they should only be held across *short,
On Thu, 28 Dec 2023 at 09:27, jian he wrote:
>
> On Wed, Dec 20, 2023 at 8:27 PM Masahiko Sawada wrote:
> >
> >
> > Why do we need to use SPI? I think we can form heap tuples and insert
> > them to the error table. Creating the error table also doesn't need to
> > use SPI.
> >
> Thanks for
> On Jan 2, 2024, at 19:35, Aleksander Alekseev
> wrote:
>
> Thanks for the updated patch.
>
> cfbot seems to have some complaints regarding compiler warnings and
> also building the patch on Windows:
>
> http://cfbot.cputube.org/
Thanks for the information. Here is the updated patch.
On Thu, Jan 4, 2024, at 3:05 AM, Amit Kapila wrote:
> Won't it be a better user experience that after setting up the target
> server as a logical replica (subscriber), it started to work
> seamlessly without user intervention?
If we have an option to control the replication slot removal (default
On Thu, Jan 4, 2024, at 2:41 AM, Amit Kapila wrote:
> So, you also seem to be saying that it is not required once
> pg_subscriber has promoted it. So, why it should be optional to remove
> physical_replication_slot? I think we must remove it from the primary
> unless there is some other reason.
Hi,
> This approach avoids the need to rewrite SQL or give special meaning to SQL
> comments.
SQLCommenter already has a good amount of support and is referenced in
opentelemetry https://github.com/open-telemetry/opentelemetry-sqlcommenter
So the goal was more to leverage the existing trace
Robert Haas writes:
> I'm not sure that the approach this patch takes is correct in detail,
> but I kind of agree with you about the overall point. I mean, the idea
> of the PANIC is to avoid having the system just sit there in a state
> from which it will never recover ... but it can also have
On Wed, Jan 3, 2024 at 6:36 PM Nico Williams wrote:
> On Tue, Jan 02, 2024 at 10:14:16AM -0500, Robert Haas wrote:
> > It seems like a pretty significant savings no matter what. Suppose the
> > backup_manifest file is 2GB, and instead of creating a 2GB buffer, you
> > create an 1MB buffer and
Hi,
Here's a somewhat reworked version of the patch. My initial goal was to
see if it could adopt the StreamingRead API proposed in [1], but that
turned out to be less straight-forward than I hoped, for two reasons:
(1) The StreamingRead API seems to be designed for pages, but the index
code
Hi Matthias and Robert,
Matthias van de Meent writes:
> On Thu, 4 Jan 2024 at 08:09, Andy Fan wrote:
>>
>> My question is if someone doesn't obey the rule by mistake (everyone
>> can make mistake), shall we PANIC on a production environment? IMO I
>> think it can be a WARNING on a production
Hi,
On Thu, Jan 04, 2024 at 10:27:31AM +0530, shveta malik wrote:
> On Thu, Jan 4, 2024 at 9:18 AM shveta malik wrote:
> >
> > On Wed, Jan 3, 2024 at 6:33 PM Zhijie Hou (Fujitsu)
> > wrote:
> > >
> > > On Tuesday, January 2, 2024 6:32 PM shveta malik
> > > wrote:
> > > > On Fri, Dec 29, 2023
On Thu, Jan 4, 2024 at 2:09 AM Andy Fan wrote:
> My question is if someone doesn't obey the rule by mistake (everyone
> can make mistake), shall we PANIC on a production environment? IMO I
> think it can be a WARNING on a production environment and be a stuck
> when 'ifdef USE_ASSERT_CHECKING'.
>
Hello Tom,
04.01.2024 02:39, Tom Lane wrote:
Buildfarm member skink has failed 3 times in
035_standby_logical_decoding.pl in the last couple of days:
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=skink=2024-01-03%2020%3A07%3A15
On Tue, Jan 2, 2024 at 3:36 PM Ashutosh Bapat
wrote:
>
> I will look at 0002 next.
One more comment on 0001
InjectionPointAttach() doesn't test whether the given function exists
in the given library. Even if InjectionPointAttach() succeeds,
INJECTION_POINT might throw error because the function
Hello!
> Correct, but there are changes being discussed where we would freeze
> tuples during pruning as well [0], which would invalidate that
> implementation detail. And, if I had to choose between improved
> opportunistic freezing and improved R/CIC, I'd probably choose
> improved freezing
(typo in the subject fixed)
> In the following paragraph in information_schema:
>
> character encoding form
>
>
>An encoding of some character repertoire. Most older character
>repertoires only use one encoding form, and so there are no
>separate names for
On Thu, Jan 4, 2024 at 5:23 AM Michael Paquier wrote:
>
> > 0003 and 0004 add tests to the test_injection_points module. Is the idea
> > that we'd add any tests that required injection points here? I think it'd
> > be better if we could move the tests closer to the logic they're testing,
> >
On Thu, Jan 4, 2024 at 4:34 PM Amit Kapila wrote:
> > But what good is having the same publications as primary
> > also on logical replica?
> >
>
> The one use case that comes to my mind is to set up bi-directional
> replication. The publishers want to subscribe to the new subscriber.
Hmm. Looks
Hello Amit,
03.01.2024 14:42, Amit Kapila wrote:
So I started to think about other approach: to perform unlink as it's
implemented now, but then wait until the DELETE_PENDING state is gone.
There is a comment in the code which suggests we shouldn't wait
indefinitely. See "However, we won't
On Mon, 25 Dec 2023 at 15:12, Michail Nikolaev
wrote:
>
> Hello!
>
> It seems like the idea of "old" snapshot is still a valid one.
>
> > Should this deal with any potential XID wraparound, too?
>
> As far as I understand in our case, we are not affected by this in any way.
> Vacuum in our table
On Thu, Jan 4, 2024 at 12:22 PM Shlok Kyal wrote:
>
> Hi,
> I was testing the patch with following test cases:
>
> Test 1 :
> - Create a 'primary' node
> - Setup physical replica using pg_basebackup "./pg_basebackup –h
> localhost –X stream –v –R –W –D ../standby "
> - Insert data before and
On Thu, Jan 4, 2024 at 12:30 PM Ashutosh Bapat
wrote:
>
> On Wed, Jan 3, 2024 at 2:49 PM Amit Kapila wrote:
> >
> > c) Drop the replication slots d) Drop the
> > > publications
> > >
> >
> > I am not so sure about dropping publications because, unlike
> > subscriptions which can start to pull
On Thu, 4 Jan 2024 at 08:09, Andy Fan wrote:
>
> My question is if someone doesn't obey the rule by mistake (everyone
> can make mistake), shall we PANIC on a production environment? IMO I
> think it can be a WARNING on a production environment and be a stuck
> when 'ifdef USE_ASSERT_CHECKING'.
>
On Tue, 2 Jan 2024 at 15:58, Amit Kapila wrote:
>
> On Fri, Dec 29, 2023 at 2:26 PM vignesh C wrote:
> >
> > On Thu, 28 Dec 2023 at 15:59, Amit Kapila wrote:
> > >
> > > On Wed, Dec 13, 2023 at 12:09 PM vignesh C wrote:
> > > >
> > > > Thanks for the comments, the attached v25 version patch
On Wed, 3 Jan 2024 at 11:25, Amit Kapila wrote:
>
> On Wed, Jan 3, 2024 at 6:21 AM Michael Paquier wrote:
> >
> > On Tue, Jan 02, 2024 at 03:58:25PM +0530, Amit Kapila wrote:
> > > On Fri, Dec 29, 2023 at 2:26 PM vignesh C wrote:
> > >> Thanks, the changes look good.
> > >
> > > Pushed.
> >
> >
On Wed, Jan 3, 2024 at 4:57 PM Bertrand Drouvot
wrote:
>
> On Wed, Jan 03, 2024 at 04:20:03PM +0530, Amit Kapila wrote:
> > On Fri, Dec 29, 2023 at 12:32 PM Amit Kapila
> > wrote:
> > >
> > > On Fri, Dec 29, 2023 at 6:59 AM Masahiko Sawada
> > > wrote:
> > > >
> > > > On Wed, Dec 27, 2023 at
Hi,
We have documentation on how to upgrade "publisher" and "subscriber"
at [1], but currently we do not have any documentation on how to
upgrade logical replication clusters.
Here is a patch to document how to upgrade different logical
replication clusters: a) Upgrade 2 node logical replication
> On 4 Jan 2024, at 07:14, Japin Li wrote:
>
> Does the timeout is too short for testing? I see the timeouts for lock_timeout
> and statement_timeout is more bigger than transaction_timeout.
Makes sense. Done. I've also put some effort into fine-tuning timeouts Nik's
case tests. To have
78 matches
Mail list logo
