Re: [HACKERS] RTE_NAMEDTUPLESTORE, enrtuples and comments

On Sun, Jun 11, 2017 at 6:25 PM, Noah Misch wrote: > While completing my annual src/backend/nodes/*funcs.c audit, I noticed defects > in commit 18ce3a4 changes to RangeTblEntry: > > 1. Field relid is under a comment saying it is valid for RTE_RELATION only. The comment is out of date. Here's a f

[HACKERS] PG10 Partitioned tables and relation_is_updatable()

Hi, It looks like relation_is_updatable() didn't get the message about partitioned tables. Thus, for example, information_schema.views and information_schema.columns report that simple views built on top of partitioned tables are non-updatable, which is wrong. Attached is a patch to fix this. I t

[HACKERS] Buildfarm failures on woodlouse (in ecpg-check)

Hello, my buildfarm animal woodlouse (Visual Studio 2013 on Windows 7) stopped working correctly some months ago. After Tom kindly prodded me into fixing it, I noticed that I had configured it to skip the ecpg-check step because one of the tests in the "thread" section (not always the same) n

[HACKERS] Re: [BUGS] BUG #14682: row level security not work with partitioned table

On 06/09/2017 02:52 PM, Joe Conway wrote: > On 06/09/2017 06:16 AM, Joe Conway wrote: >> On 06/08/2017 11:09 PM, Noah Misch wrote: >>> On Wed, Jun 07, 2017 at 08:45:20AM -0700, Joe Conway wrote: On 06/07/2017 06:49 AM, Mike Palmiotto wrote: > I ended up narrowing it down to 4 tables (one

Re: [HACKERS] PG10 Partitioned tables and relation_is_updatable()

On 06/11/2017 04:32 AM, Dean Rasheed wrote: > It looks like relation_is_updatable() didn't get the message about > partitioned tables. Thus, for example, information_schema.views and > information_schema.columns report that simple views built on top of > partitioned tables are non-updatable, which

Re: [HACKERS] PG10 Partitioned tables and relation_is_updatable()

On 06/11/2017 08:55 AM, Joe Conway wrote: > On 06/11/2017 04:32 AM, Dean Rasheed wrote: >> It looks like relation_is_updatable() didn't get the message about >> partitioned tables. Thus, for example, information_schema.views and >> information_schema.columns report that simple views built on top of

[HACKERS] regproc and when to schema-qualify

I was idly following along in GSoC 2017: Foreign Key Arrays when I noticed this: =# select * from pg_amproc where amprocfamily = 2745; amprocfamily | amproclefttype | amprocrighttype | amprocnum | amproc --++-+---+ 2745 | 2277

Re: [HACKERS] Make ANALYZE more selective about what is a "most common value"?

On 05/06/17 09:30, Tom Lane wrote: > First, I think we need a larger hard floor on the number of occurrences > of a value that're required to make ANALYZE decide it is a "most common > value"... > > Second, the code also has a rule that potential MCVs need to have an > estimated frequency at least

Re: [HACKERS] regproc and when to schema-qualify

Chapman Flack writes: > The manual says regproc "will display schema-qualified names on output > if the object would not be found in the current search path without > being qualified." That's less than the full truth :-( > Is regproc displaying the schema in this case because there are two > ove

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

On Thu, Jun 1, 2017 at 8:40 AM, Tom Lane wrote: > The thing that would actually have a chance of improving matters for Q20 > would be if we could see our way to looking through the aggregation > subquery and applying the foreign key constraint for lineitem. That > seems like a research project th

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

Peter Geoghegan writes: > On Thu, Jun 1, 2017 at 8:40 AM, Tom Lane wrote: >> The thing that would actually have a chance of improving matters for Q20 >> would be if we could see our way to looking through the aggregation >> subquery and applying the foreign key constraint for lineitem. That >> s

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

On Sun, Jun 11, 2017 at 10:36 AM, Tom Lane wrote: >> Do you mean teaching the optimizer to do something like this?: > > Uh, no. I don't think we want to add any run-time checks. The point in > this example is that we'd get a better rowcount estimate if we noticed > that the FK constraint could b

Re: [HACKERS] PostgreSQL 10 changes in exclusion constraints - did something change? CASE WHEN behavior oddity

Alvaro Herrera writes: > Interesting stuff. Here's a small recommendation for a couple of those > new messages. Hm. I don't object to folding those two messages into one, but now that I look at it, the text needs some more work anyway, perhaps. What we're actually checking is not so much wheth

Re: [HACKERS] Broken hint bits (freeze)

> 8 июня 2017 г., в 17:03, Amit Kapila написал(а): > > On Thu, Jun 8, 2017 at 6:49 PM, Dmitriy Sarafannikov > wrote: >> >>> Why didn't rsync made the copies on master and replica same? >> >> Because rsync was running with —size-only flag. >> > > IIUC the situation, the new WAL and updated p

Re: [HACKERS] PG10 Partitioned tables and relation_is_updatable()

On 11 June 2017 at 16:59, Joe Conway wrote: > On 06/11/2017 08:55 AM, Joe Conway wrote: >> Yeah, I noticed the same while working on the RLS related patch. I did >> not see anything else in rewriteHandler.c but it is probably worth >> looking wider for other omissions. > > So in particular: > > #d

Re: [HACKERS] Make ANALYZE more selective about what is a "most common value"?

Dean Rasheed writes: > I think we should attempt to come up with a more principled approach > to this, taking into account the table and sample sizes. Here's what I > found, after a bit of research: Thanks for doing some legwork on this! > A common initial rule of thumb is that the value should

Re: [HACKERS] Make ANALYZE more selective about what is a "most common value"?

On 11 June 2017 at 20:19, Tom Lane wrote: >> The standard way of doing this is to calculate the "standard error" of >> the sample proportion - see, for example [3], [4]: >> SE = sqrt(p*(1-p)/n) >> Note, however, that this formula assumes that the sample size n is >> small compared to the populat

Re: [HACKERS] Buildfarm failures on woodlouse (in ecpg-check)

On 06/11/2017 11:33 AM, Christian Ullrich wrote: > Hello, > > my buildfarm animal woodlouse (Visual Studio 2013 on Windows 7) > stopped working correctly some months ago. After Tom kindly prodded me > into fixing it, I noticed that I had configured it to skip the > ecpg-check step because one of

Re: pgindent (was Re: [HACKERS] [COMMITTERS] pgsql: Preventive maintenance in advance of pgindent run.)

>>> * the comments get formatted differently for -ts4 than -ts8 Still haven't put any thought into it, so I still don't know what to do here. >>> * extra spacing getting inserted for fairly long labels I think the fix is as easy as not producing the space. I committed that. >>> * some enum decl

[HACKERS] Transactional sequence stuff breaks pg_upgrade

I believe I've identified the reason why skink and some other buildfarm members have been failing the pg_upgrade test recently. It is that recent changes in sequence support have caused binary-upgrade restore runs to do some sequence OID/relfilenode assignments without any heed to the OIDs that pg

Re: [HACKERS] PG10 transition tables, wCTEs and multiple operations on the same table

> "Robert" == Robert Haas writes: Robert> I don't see a reason why MakeTransitionCaptureState needs to Robert> force the tuplestores into TopTransactionContext or make them Robert> owned by TopTransactionResourceOwner. Nor do I, and I'm pretty sure it's leaking memory wholesale within a t

Re: [HACKERS] SCRAM in the PG 10 release notes

On 11/05/17 09:20, Heikki Linnakangas wrote: On 05/11/2017 07:03 AM, Michael Paquier wrote: On Thu, May 11, 2017 at 11:50 AM, Bruce Momjian wrote: I have added this as an open item because we will have to wait to see where we are with driver support as the release gets closer. As Postgres

Re: [HACKERS] transition table behavior with inheritance appears broken

> "Andrew" == Andrew Gierth writes: Andrew> I have it; I will post a status update before 23:59 BST on 11 Andrew> Jun. This is that status update. I am still studying Thomas' latest patch set; as I mentioned in another message, I've confirmed a memory leak, and I expect further work may b

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

On Sun, Jun 11, 2017 at 10:27 AM, Peter Geoghegan wrote: > Note that I introduced a new, redundant exists() in the agg_lineitem > fact table subquery. It now takes 23 seconds for me on Tomas' 10GB > TPC-H dataset, whereas the original query took over 90 minutes. > Clearly we're missing a trick or

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

Hi, On 6/11/17 7:54 PM, Peter Geoghegan wrote: On Sun, Jun 11, 2017 at 10:36 AM, Tom Lane wrote: Do you mean teaching the optimizer to do something like this?: Uh, no. I don't think we want to add any run-time checks. The point in this example is that we'd get a better rowcount estimate if

Re: [HACKERS] TPC-H Q20 from 1 hour to 19 hours!

On Sun, Jun 11, 2017 at 4:10 PM, Tomas Vondra wrote: > I do strongly recommend reading this paper analyzing choke points of > individual TPC-H queries: > > http://oai.cwi.nl/oai/asset/21424/21424B.pdf > > It's slightly orthogonal to the issue at hand (poor estimate in Q20 causing > choice of i

Re: [HACKERS] Transactional sequence stuff breaks pg_upgrade

I wrote: > I believe I've identified the reason why skink and some other buildfarm > members have been failing the pg_upgrade test recently. > ... > Not sure what we want to do about it. One idea is to make > ALTER SEQUENCE not so transactional when in binary-upgrade mode. On closer inspection, t

Re: [HACKERS] Proposal : For Auto-Prewarm.

Thanks, Amit, On Fri, Jun 9, 2017 at 8:07 PM, Amit Kapila wrote: > > Few comments on the latest patch: > + > + /* Prewarm buffer. */ > + buf = ReadBufferExtended(rel, blk->forknum, blk->blocknum, RBM_NORMAL, > + NULL); > + if (BufferIsValid(buf)) > + ReleaseBuffer(buf); > + > + old_blk = blk; >

[HACKERS] documentation typo in ALTER TABLE example

Hi, Attached is a simple patch to fix a documentation typo in the ALTER TABLE example. Regards, -- Yugo Nagata diff --git a/doc/src/sgml/ref/alter_table.sgml b/doc/src/sgml/ref/alter_table.sgml index 56ea830..4c61c44 100644 --- a/doc/src/sgml/ref/alter_table.sgml +++ b/doc/src/sgml/ref/alter_t

Re: [HACKERS] Re: Alter subscription..SET - NOTICE message is coming for table which is already removed

On Fri, Jun 9, 2017 at 10:50 PM, Peter Eisentraut wrote: > On 5/30/17 13:25, Masahiko Sawada wrote: >> However there is one more problem here; if the relation status entry >> is deleted while corresponding table sync worker is waiting to be >> changed its status, the table sync worker can be orpha

Re: [HACKERS] documentation typo in ALTER TABLE example

> Hi, > > Attached is a simple patch to fix a documentation typo in > the ALTER TABLE example. Or the original author's intention might have been something like this: --- a/doc/src/sgml/ref/alter_table.sgml +++ b/doc/src/sgml/ref/alter_table.sgml @@ -1399,7 +1399,7 @@ ALTER TABLE cities Deta

Re: [HACKERS] transition table behavior with inheritance appears broken (was: Declarative partitioning - another take)

On Sat, Jun 10, 2017 at 6:08 AM, Robert Haas wrote: > I have spent some time now studying this patch. I might be missing > something, but to me this appears to be in great shape. A few minor > nitpicks: > > -if ((event == TRIGGER_EVENT_DELETE && > !trigdesc->trig_delete_after_row) || > -

Re: [HACKERS] PG10 transition tables, wCTEs and multiple operations on the same table

On Mon, Jun 12, 2017 at 9:29 AM, Andrew Gierth wrote: >> "Robert" == Robert Haas writes: > > Robert> I don't see a reason why MakeTransitionCaptureState needs to > Robert> force the tuplestores into TopTransactionContext or make them > Robert> owned by TopTransactionResourceOwner. > > Nor

Re: [HACKERS] Transition tables vs ON CONFLICT

On Fri, Jun 9, 2017 at 4:10 PM, Thomas Munro wrote: > On Wed, Jun 7, 2017 at 10:47 AM, Peter Geoghegan wrote: >> I suppose you'll need two tuplestores for the ON CONFLICT DO UPDATE >> case -- one for updated tuples, and the other for inserted tuples. > > [...] > > Here is a patch implementing the

Re: [HACKERS] Adding support for Default partition in partitioning

Hi Ashutosh, I tried to look into your refactoring code. When applied all 3 patches, I got some regression failures, I have fixed all of them now in attached patches, attached the regression.diffs. Moving further, I have also made following changes in attached patches: *1. 0001-Refactor-ATExecAt

Re: [HACKERS] RTE_NAMEDTUPLESTORE, enrtuples and comments

On Sun, Jun 11, 2017 at 11:11 PM, Thomas Munro wrote: > On Sun, Jun 11, 2017 at 6:25 PM, Noah Misch wrote: >> This fourth point is not necessarily a defect: I wonder if RangeTblEntry is >> the right place for enrtuples. It's a concept regularly seen in planner data >> structures but not otherwis

Re: [HACKERS] Adding support for Default partition in partitioning

While the refactoring seems a reasonable way to re-use existing code, that may change based on the discussion in [1]. Till then please keep the refactoring patches separate from the main patch. In the final version, I think the refactoring changes to ATAttachPartition and the default partition supp

Re: [HACKERS] ECPG: WHENEVER statement with DO CONTINUE action

Hi, On 2017/06/10 12:23, Vinayak Pokale wrote: Thank you for your reply On Jun 9, 2017 5:39 PM, "Michael Meskes" > wrote: > > Could you please add a "DO CONTINUE" case to one of the test cases? Or > add a new one? We would need a test case IMO. > Yes I will add t

Re: [HACKERS] Getting server crash on Windows when using ICU collation

PFA patch that fixes the issue described in above thread. As mentioned in the above thread, the crash is basically happening in varstr_cmp() function and it's only happening on Windows because in varstr_cmp(), if the collation provider is ICU, we don't even think of calling ICU functions to compa

Re: [HACKERS] documentation typo in ALTER TABLE example

On 2017/06/12 11:01, Tatsuo Ishii wrote: >> Hi, >> >> Attached is a simple patch to fix a documentation typo in >> the ALTER TABLE example. > > Or the original author's intention might have been something like > this: > > --- a/doc/src/sgml/ref/alter_table.sgml > +++ b/doc/src/sgml/ref/alter_tabl

Re: [HACKERS] Buildfarm failures on woodlouse (in ecpg-check)

* Andrew Dunstan wrote: On 06/11/2017 11:33 AM, Christian Ullrich wrote: To build correctly, it requires defining _WIN32_WINNT to be 0x600 or above (and using an SDK that knows about InitOnceExecuteOnce()). We already define _WIN32_WINNT to be 0x0600 on all appropriate platforms (Vista/200

Re: [HACKERS] documentation typo in ALTER TABLE example

>> Amit Langote, which one was your intention? > > I wanted to show DETACH PARTITION command's usage with a range partitioned > table (detaching the "oldest" partition). > > So, we should apply Nagata-san's patch. Thank you for the confirmation. I have pushed the patch. Best regards, -- Tatsuo

[HACKERS] Why forcing Hot_standby_feedback to be enabled when creating a logical decoding slot on standby

Hi, I have created a logical decoding slot on a standby but i haven't enabled Hot_standby_feedback.What are the test cases where this setup will fail? Thanks, Sanyam Jain

Re: [HACKERS] Adding support for Default partition in partitioning

On Mon, Jun 12, 2017 at 9:39 AM, Ashutosh Bapat < ashutosh.ba...@enterprisedb.com> wrote: > While the refactoring seems a reasonable way to re-use existing code, > that may change based on the discussion in [1]. Till then please keep > the refactoring patches separate from the main patch. In the f

Re: [HACKERS] ICU support on Windows

On Sat, Jun 10, 2017 at 6:40 PM, Ashutosh Sharma wrote: > Currently, we cannot perform ICU enabled build for postgres on Windows > platform. However, this can be done on Linux platforms using > '--with-icu' configuration parameter. Attached is the patch that > allows us to perform icu enabled buil

Re: [HACKERS] Why forcing Hot_standby_feedback to be enabled when creating a logical decoding slot on standby

On Mon, Jun 12, 2017 at 3:16 PM, sanyam jain wrote: > I have created a logical decoding slot on a standby but i haven't enabled > Hot_standby_feedback.What are the test cases where this setup will fail? hot_standby_feedback needs to be enabled at all times in logical decoding so as the node does