Re: [HACKERS] Walsender timeouts and large transactions

Ouch.. I'd doubly mistaked. > I found that the patch is the latest one and will look this > soon. Sorry for the ignorance. Thats...wrong. Sorry. There's no new patch since the Reboer's comment. I think this is just a bug fix and needs no more argument on its functionality. (and might ought to

Speed up the removal of WAL files

Hello, The attached patch speeds up the removal of WAL files in the old timelines. I'll add this to the next CF. BACKGROUND == We need to meet a severe availability requirement of a potential customer. They will use synchronous streaming

Re: Add PGDLLIMPORT lines to some variables

Hi 2017-11-16 23:59 GMT+01:00 Brian Cloutier : > Hello hackers, > > I'm porting Citus to Windows and found that we use some variables which PG > doesn't export; here is a patch which adds PGDLLIMPORT declarations to > those variables. This is unfortunately required on

Re: [HACKERS] Partition-wise aggregation/grouping

On Thu, Nov 16, 2017 at 6:02 AM, David Rowley wrote: > On 16 November 2017 at 05:57, Konstantin Knizhnik > wrote: >> The main problem IMHO is that there are a lot of different threads and >> patches related with this topic:( >> And it is

default range partition and constraint exclusion

Hi. While working on the patch for partition pruning for declarative partitioned tables, I noticed that default range partition will fail to be included in a plan in certain cases due to pruning by constraint exclusion. Consider a multi-column range-partitioned table: create table mc2p (a int,

Re: [HACKERS] [PATCH] A hook for session start

On Fri, Nov 17, 2017 at 9:08 AM, Fabrízio de Royes Mello wrote: > Hi all, > > Attached new version of the patch fixing issues about installcheck and some > assertions reported before (based on Michael Paquier code): > The assertion failure which i reported earlier -[1]

Re: [HACKERS] Walsender timeouts and large transactions

Hello, At Fri, 17 Nov 2017 13:24:08 +0900 (Tokyo Standard Time), Kyotaro HORIGUCHI wrote in <20171117.132408.85564852.horiguchi.kyot...@lab.ntt.co.jp> > > Well, even the CHECK_FOR_INTERRUPTS() can be called unconditionally yes. > > It just seems like it's

Re: [HACKERS] Walsender timeouts and large transactions

Hello, At Fri, 3 Nov 2017 15:54:09 +0100, Petr Jelinek wrote in > Hi, > > thanks for checking. > > On 02/11/17 10:00, Robert Haas wrote: > > On Wed, Nov 1, 2017 at 8:19 PM, Petr Jelinek > >

Re: [HACKERS] Improve catcache/syscache performance.

Hello, At Fri, 13 Oct 2017 14:36:45 -0700, Andres Freund wrote in <20171013213645.fuolk62xaedj3...@alap3.anarazel.de> > Pushed after making the adaptions you suggested, pgindenting, and fixing > one bug (cstring lookups + datumCopy() = not good). > > Greetings, > > Andres

Re: _WINSOCK_DEPRECATED_NO_WARNINGS

Glen Knowles writes: > On Thu, Nov 16, 2017 at 8:37 AM, Tom Lane wrote: >> but evidently it chose the wrong cutoff for when to enable that >> symbol, because woodlouse is (or claims to be) running VS2013. > It's actually checking the wrong thing, the

Re: [HACKERS] Assertion failure when the non-exclusive pg_stop_backup aborted.

On Thu, Nov 16, 2017 at 8:17 PM, Masahiko Sawada wrote: > On Thu, Nov 16, 2017 at 1:11 PM, Michael Paquier > wrote: >> + /* >> +* Quick exit if session is not keeping around a non-exclusive backup >> +* already started. >> +*/ >> +

Re: [HACKERS] Transaction control in procedures

On Thu, Nov 16, 2017 at 5:35 PM, Simon Riggs wrote: > On 14 November 2017 at 13:09, Peter Eisentraut > wrote: > >>> *) Will pg_cancel_backend() cancel the currently executing statement >>> or the procedure? (I guess probably the procedure

Re: [HACKERS] Parallel Hash take II

Hi, On 2017-11-15 14:09:13 -0500, Robert Haas wrote: > On Wed, Nov 15, 2017 at 1:35 PM, Andres Freund wrote: > > But this does bug me, and I think it's what made me pause here to make a > > bad joke. The way that parallelism treats work_mem makes it even more > > useless of

Re: [HACKERS] Transaction control in procedures

On 14 November 2017 at 13:09, Peter Eisentraut wrote: >> *) Will pg_cancel_backend() cancel the currently executing statement >> or the procedure? (I guess probably the procedure but I'm curious) > > Same as the way it currently works. It will raise an

Re: [HACKERS] Transaction control in procedures

On 15 November 2017 at 16:36, Peter Eisentraut wrote: > On 11/8/17 18:48, Simon Riggs wrote: >> What would happen if some of the INSERTs failed? Where would control >> go to? (Maybe this is just "no change" in this particular proc) > > An exception is raised and

Re: pspg - psql pager

On Thu, Nov 16, 2017 at 11:04 AM, Pavel Stehule wrote: > 2017-11-16 16:40 GMT+01:00 Aaron W. Swenson : >> I likes it. I likes it a lot. >> >> And, you’ve got a PR on GitHub from me. >> >> It’s available on Gentoo now! This would be great package

Re: Transaction control in procedures

On Thu, Nov 16, 2017 at 12:36 PM, Peter Eisentraut wrote: > On 11/16/17 07:04, legrand legrand wrote: >> We are just opening the "close cursors on/at commit" specification ;o) >> >> - MS SQL server: cursor_close_on_commit >> - Firebird: close_cursors_at_commit

Add PGDLLIMPORT lines to some variables

Hello hackers, I'm porting Citus to Windows and found that we use some variables which PG doesn't export; here is a patch which adds PGDLLIMPORT declarations to those variables. This is unfortunately required on Windows for extensions to be able to use those variables, and appears to already have

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Thu, Nov 16, 2017 at 2:16 PM, Jeff Janes wrote: > The only reference to super-exclusive lock in src/backend/access/gin/README, > that I can find, is about posting trees, not pending lists. Can you quote > or give line numbers of the section you are referring to? That's

Re: [HACKERS] SERIALIZABLE on standby servers

On 7 November 2016 at 23:56, Thomas Munro wrote: > The patch works by teaching the standby how to do somethings similar > to what SERIALIZABLE READ ONLY DEFERRABLE does on the primary server, > with some help from the primary server in the form of extra >

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Mon, Nov 13, 2017 at 8:08 PM, Peter Geoghegan wrote: > On Mon, Nov 13, 2017 at 6:56 PM, Masahiko Sawada > wrote: > >/* > > * We would like to prevent concurrent cleanup process. For that we > will > > * lock metapage in exclusive mode using

Re: [JDBC] [HACKERS] Channel binding support for SCRAM-SHA-256

On Thu, Nov 16, 2017 at 10:47 PM, Peter Eisentraut wrote: > Pushed 0001, will continue with 0002. Thanks! -- Michael

Re: bgwriter_lru_maxpages range in postgresql.conf

On Thu, Nov 16, 2017 at 11:10 AM, Robert Haas wrote: > On Thu, Nov 16, 2017 at 1:26 PM, Jeff Janes wrote: > > (My earlier post has been stuck in infinite moderation due to the use of > a > > common 6 letter abbreviation for configuration, so I will

Re: [HACKERS] [PATCH] Generic type subscripting

> On 16 November 2017 at 12:40, Arthur Zakirov wrote: > > Actually it is not only way to return isnull information. What i've meant is that it's the only way if we want to keep the function signature. Actually I would prefer this (container, internal) -> extracted

Re: Schedule for migration to pglister

On Thu, Nov 16, 2017 at 4:51 PM, Justin Clift wrote: > On 2017-11-16 15:34, Stephen Frost wrote: > >> Magnus, all, >> >> * Magnus Hagander (mag...@hagander.net) wrote: >> >>> mbox access should now be restored, and as Stephen says, now handled by >>> the >>> Date in the

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Thu, Nov 16, 2017 at 12:49 PM, Peter Geoghegan wrote: > But there is only ever one page locked, the meta-page. And, it's > always an ExclusiveLock. I don't see much use for deadlock avoidance. I meant deadlock detection. -- Peter Geoghegan

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Mon, Nov 13, 2017 at 8:48 PM, Peter Geoghegan wrote: > One thing that really bothers me about commit e2c79e14 is that > LockPage() is called, not LockBuffer(). GIN had no LockPage() calls > before that commit, and is now the only code in the entire system that > calls

Re: Inlining functions with "expensive" parameters

Hi, On 2017-11-16 15:27:59 -0500, Tom Lane wrote: > Andres Freund writes: > > On November 16, 2017 11:44:52 AM PST, Tom Lane wrote: > >> Yeah, there's no mechanism like that now, but there could be. > > > Right, but it doesn't sound that hard to

Re: [HACKERS] Remove duplicate setting in test/recovery/Makefile

On Thu, Nov 16, 2017 at 7:15 AM, Masahiko Sawada wrote: > Thank you. I've added this to the next CF so as not to forget. Committed. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Thu, Nov 16, 2017 at 12:29 PM, Robert Haas wrote: > I am also clear on what the consequences of this bug are. It seems > like it could harm insert performance by making us wait when we > shouldn't, but can it cause corruption? That I'm not sure about. If > there's

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Thu, Nov 16, 2017 at 7:08 AM, Masahiko Sawada wrote: > Agreed, that's better. Attached updated patch. > Also I've added this to the next CF so as not to forget. Committed and back-patched. While I'm fairly sure this is a correct fix, post-commit review from someone who

Re: Inlining functions with "expensive" parameters

Andres Freund writes: > On November 16, 2017 11:44:52 AM PST, Tom Lane wrote: >> Yeah, there's no mechanism like that now, but there could be. > Right, but it doesn't sound that hard to introduce. Basically there'd need to > be a WithParamValue node,

Re: Transaction control in procedures

On 11/16/17 07:04, legrand legrand wrote: > We are just opening the  "close cursors on/at commit" specification ;o) > > - MS SQL server: cursor_close_on_commit > - Firebird: close_cursors_at_commit > - DB2: "with hold" syntax > - ... > > I think it a plus to support keeping opened cursors at

bgwriter_lru_maxpages range in postgresql.conf

(My earlier post has been stuck in infinite moderation due to the use of a common 6 letter abbreviation for configuration, so I will try again without it) With v10, commit 14ca9abfbe4643408a, the upper limit on bgwriter_lru_maxpages was changed from 1000 to INT_MAX / 2, but the

Re: Inlining functions with "expensive" parameters

Hi, On 2017-11-16 09:37:29 -0800, Paul Ramsey wrote: > On Thu, Nov 9, 2017 at 12:11 PM, Paul Ramsey > wrote: > > > All, > > > > As we try and make PostGIS more "parallel sensitive" we have been added > > costs to our functions, so that their relative CPU cost is more

Re: [HACKERS] Inconsistencies between pg_settings and postgresql.conf

On Thu, Nov 16, 2017 at 6:08 AM, Adrian Escoms wrote: > It makes sense, so I attach the new diff without these lines. Thanks. Committed. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company

GetTupleForTrigger could use a header comment

GetTupleForTrigger seems to be a fairly important part of the trigger machinery, yet it has no header comment. I think it should, although I don't understand it well enough at this point to write one. Right now, one is left to guess at the overall purpose of the function and the meanings of the

Re: pgsql: Disable installcheck tests for test_session_hooks

On Thu, Nov 16, 2017 at 2:21 PM, Andrew Dunstan < andrew.duns...@2ndquadrant.com> wrote: > > > > On 11/16/2017 09:14 AM, Ashutosh Sharma wrote: > >> > >> Just sending an updated patch to cope with the fact that this module > >> does not need to define the EXTENSION and DATA fields, and does not >

Re: pgsql: Add hooks for session start and session end

On Thu, Nov 16, 2017 at 3:00 PM, Fabrízio de Royes Mello < fabriziome...@gmail.com> wrote: > > > > On Wed, Nov 15, 2017 at 11:26 PM, Andrew Dunstan < andrew.duns...@2ndquadrant.com> wrote: > > > > > > On 11/15/2017 05:07 PM, Michael Paquier wrote: > > > On Thu, Nov 16, 2017 at 6:53 AM, Andrew

Re: Treating work_mem as a shared resource (Was: Parallel Hash take II)

I have been pondering how to deal with work_mem for a couple of months myself and had very similar thoughts. As far as I can tell the problem goes beyond work_mem though: 1. There are several hash operations such as set-ops, hashed subplans, and hash aggregates which today are not spilling at

_WINSOCK_DEPRECATED_NO_WARNINGS

Checking the buildfarm to see if ed9b3606d broke anything, I notice that woodlouse is producing stacks of warnings like this: src/port/thread.c(134): warning C4996: 'gethostbyname': Use getaddrinfo() or GetAddrInfoW() instead or define _WINSOCK_DEPRECATED_NO_WARNINGS to disable deprecated API

Re: pgsql: Disable installcheck tests for test_session_hooks

On 11/16/2017 09:14 AM, Ashutosh Sharma wrote: >> >> Just sending an updated patch to cope with the fact that this module >> does not need to define the EXTENSION and DATA fields, and does not >> need a sql and a control file.. > While running make check with 'session_hook_makefile_v2.patch' >

Re: [HACKERS] Issues with logical replication

On Wed, Nov 15, 2017 at 8:20 PM, Stas Kelvich wrote: > I did a sketch of first approach just to confirm that it solves the problem. > But there I hold ProcArrayLock during update of flag. Since only reader is > GetRunningTransactionData it possible to have a custom lock

Re: Schedule for migration to pglister

Magnus, all, * Magnus Hagander (mag...@hagander.net) wrote: > mbox access should now be restored, and as Stephen says, now handled by the > Date in the message. Let me know if there are any issues with the new ones. > (They're now generated by the same code as the per-thread ones). And just fyi,

pgMail 1.4 Released!

Greetings all! Just a quick note to let you know I have updated and released a new version of pgMail for the community. The last time I posted to this list was in 2001, but not too much has changed! This little package gets tons of downloads and seems to be very helpful for those who want to

Re: [COMMITTERS] pgsql: Add hash partitioning.

Thanks, Michael & Robert for your suggestions, and apologize for delayed response On Tue, Nov 14, 2017 at 6:02 AM, Michael Paquier wrote: > On Tue, Nov 14, 2017 at 3:59 AM, Robert Haas wrote: >> On Mon, Nov 13, 2017 at 3:24 AM, amul sul

Re: pgsql: Disable installcheck tests for test_session_hooks

Hi Michael, On Thu, Nov 16, 2017 at 6:44 PM, Michael Paquier wrote: > > On Thu, Nov 16, 2017 at 3:45 PM, Michael Paquier > wrote: > > On Thu, Nov 16, 2017 at 11:41 AM, Tom Lane wrote: > >> Andrew Dunstan

Re: [HACKERS] Transaction control in procedures

On Wed, Nov 15, 2017 at 3:42 PM, Peter Eisentraut wrote: > On 11/15/17 09:54, Merlin Moncure wrote: >> ... I noticed that: >> *) now() did not advance with commit and, >> *) xact_start via pg_stat_activity did not advance >> >> Shouldn't both of those advance

Re: [JDBC] [HACKERS] Channel binding support for SCRAM-SHA-256

On 11/14/17 03:55, Michael Paquier wrote: > On Tue, Oct 10, 2017 at 10:12 PM, Michael Paquier > wrote: >> Attached is a new patch set with the comments from above. On top of >> that, I have changed a couple of things: >> - 0001 is unchanged, still the same refactoring

Re: test_session_hooks--1.0.sql file in 'test_session_hooks' needs correction

On Thu, Nov 16, 2017 at 7:49 PM, Ashutosh Sharma wrote: > The test_session_hooks--1.0.sql file in the newly added 'test_session_hooks' > module suggests to load the extension named 'test_hook_session' instead of > 'test_session_hooks' and i think it needs to be corrected.

Re: pgsql: Disable installcheck tests for test_session_hooks

On Thu, Nov 16, 2017 at 3:45 PM, Michael Paquier wrote: > On Thu, Nov 16, 2017 at 11:41 AM, Tom Lane wrote: >> Andrew Dunstan writes: >>> Disable installcheck tests for test_session_hooks >> >> This seems not quite right; it's

RE: Re: User defined data types in Logical Replication

Sawada-san, Thanks for your response. # And sorry again because I could not reply to your gmail # address from my environment due to security restriction. > >> We are getting the bellow error while trying use Logical Replication > >> with user defined data types in a C program (when call elog

Re: no library dependency in Makefile?

LDFLAGS in the example changed to: ''' override LDFLAGS := $(call expand_stlibs,$(STLIBS)) $(libpq_pgport) $(LDFLAGS) ''' 2017-11-16 20:50 GMT+08:00 高增琦 : > Is this a problem or not? > > > A simple fix: > 1. add a STLIBS variable in Makefiles as normal prerequisite >

Re: no library dependency in Makefile?

Is this a problem or not? A simple fix: 1. add a STLIBS variable in Makefiles as normal prerequisite 2. using GNU make's function to generate '-Lxxx -lxxx' for items in STLIBS For example: libpgfeutils.a in psql's Makefile:

Re: Further simplification of c.h's #include section

On 16 November 2017 at 19:14, Michael Paquier wrote: > On Thu, Nov 16, 2017 at 2:23 PM, David Rowley > wrote: >> I do get some warnings: >> >> (ClCompile target) -> >> src/test/modules/test_session_hooks/test_session_hooks.c(112):

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

On Thu, Nov 16, 2017 at 11:24 AM, Robert Haas wrote: > On Mon, Nov 13, 2017 at 3:25 AM, Masahiko Sawada > wrote: >> In ginInsertCleanup(), we lock the GIN meta page by LockPage and could >> wait for the concurrent cleaning up process if stats ==

Re: [HACKERS] A weird bit in pg_upgrade/exec.c

Tom Lane writes: Yeah, the way it is now seems outright broken. It will try to do get_bin_version on the new cluster before having done validate_exec there, violating its own comment. I think we should change this as a bug fix, independently of whatever else you had in

Re: [HACKERS] Issues with logical replication

> On 15 Nov 2017, at 23:09, Robert Haas wrote: > > Ouch. This seems like a bug that needs to be fixed, but do you think > it's related to to Petr's proposed fix to set es_output_cid? That fix > looks reasonable, since we shouldn't try to lock tuples without a > valid

Re: [HACKERS] Dynamic result sets from procedures

Peter Eisentraut wrote: > There is also one need error that needs further investigation. I've looked at this bit in the regression tests about \gexec: --- a/src/test/regress/expected/psql.out +++ b/src/test/regress/expected/psql.out @@ -232,11 +232,7 @@ union all select 'drop table

Re: [HACKERS] Assertion failure when the non-exclusive pg_stop_backup aborted.

On Thu, Nov 16, 2017 at 1:11 PM, Michael Paquier wrote: > On Thu, Nov 16, 2017 at 10:57 AM, Masahiko Sawada > wrote: >> Agreed. Attached the updated patch, please review it. Thank you for the comment. > + /* > +* Quick exit if session is

Re: [HACKERS] Inconsistencies between pg_settings and postgresql.conf

Hi, It makes sense, so I attach the new diff without these lines. Thank you, Regards Adrián Escoms On Wed, Nov 15, 2017 at 10:30 PM, Robert Haas wrote: > On Wed, Sep 13, 2017 at 11:50 AM, Adrian Escoms > wrote: > > I realized that the parameter

test_session_hooks--1.0.sql file in 'test_session_hooks' needs correction

Hi All, The test_session_hooks--1.0.sql file in the newly added 'test_session_hooks' module suggests to load the extension named 'test_hook_session' instead of 'test_session_hooks' and i think it needs to be corrected. Here are the contents available in test_session_hooks--1.0.sql file.

Re: [HACKERS] parallelize queries containing initplans

On Thu, Nov 16, 2017 at 3:34 PM, Amit Kapila wrote: > On Wed, Nov 15, 2017 at 12:25 AM, Robert Haas wrote: >> On Tue, Nov 14, 2017 at 11:00 AM, Tom Lane wrote: >>> Yeah, I'm sure it is. I have a vague recollection that there

Re: [HACKERS] Runtime Partition Pruning

On Wed, Nov 15, 2017 at 4:43 AM, David Rowley wrote: > On 15 November 2017 at 01:57, David Rowley > wrote: >> I think to do this you're going to have to store some sort of array >> that maps the partition index to the subpath in the

./configure fails for --host=i686-w64-mingw32 on Ubuntu

Hello, community. Following instructions on Wiki https://wiki.postgresql.org/wiki/Building_With_MinGW I was trying to cross compile PostgreSQL 10.1 from a Linux box (Ubuntu 16.04). But it always fails on configure step. I tried two different systems, results are the same. Log attached. I

Re: [HACKERS] [PROPOSAL] Temporal query processing with range types

2017-11-14 18:42 GMT+01:00 Tom Lane : > Robert is correct that putting this into the parser is completely the > wrong thing. If you do that, then for example views using the features > will reverse-list in the rewritten form, which we Do Not Want, even > if the rewritten form

Re: User defined data types in Logical Replication

On Wed, Nov 15, 2017 at 7:55 PM, Huong Dangminh wrote: > Hi, > >> We are getting the bellow error while trying use Logical Replication with >> user defined data types in a C program (when call elog function). >> >> ERROR: XX000: cache lookup failed for type X >>