On Thursday, August 30, 2012 7:14 PM Robert Haas wrote:
On Wed, Aug 29, 2012 at 10:25 PM, Peter Geoghegan
wrote:
> On 19 February 2012 05:24, Robert Haas wrote:
>>> I have attached tps scatterplots. The obvious conclusion appears to
>>> be that, with only 16MB of wal_buffers, the buffer "wraps a
> From: Robert Haas [mailto:robertmh...@gmail.com]
> On Thu, Aug 16, 2012 at 2:11 AM, Etsuro Fujita
> wrote:
> > Agreed. I'd like to withdraw the patch sent in the earlier post, and
propose
> to
> > update the documentation in the COPY reference page. Please find attached
> a
> > patch.
>
> I
2012/8/30 Robert Haas :
> On Mon, Aug 13, 2012 at 3:35 PM, Kohei KaiGai wrote:
>> The attached patch is a refreshed version of ALTER command
>> reworks towards the latest tree. Here is few big changes except
>> for code integration of the code to rename event triggers.
>
> This seems to have bit-r
On Thu, Jul 26, 2012 at 10:36:59AM -0400, Bruce Momjian wrote:
> > > Pg_upgrade already creates a script to analyze the cluster, so we could
> > > create another script to upgrade a standby. However, the problem with a
> > > script is that I have no idea what command people would use to do the
> >
On Thu, Aug 30, 2012 at 07:59:02PM -0700, Joe Conway wrote:
> On 08/30/2012 07:23 PM, Bruce Momjian wrote:
> > On Thu, Jul 12, 2012 at 06:01:00PM -0700, Joe Conway wrote:
> >> I'll take a look at the latter option sometime in the next few weeks and
> >> submit for the next commitfest.
> >
> > Any
Robert Haas writes:
> On Thu, Aug 30, 2012 at 4:51 PM, Tom Lane wrote:
>> Bruce Momjian writes:
>>> On Thu, May 31, 2012 at 09:20:43AM +0900, Tatsuo Ishii wrote:
Ok, I modified the part of pg_dump where tremendous number of LOCK
TABLE are issued. I replace them with single LOCK TABLE w
On Thursday, August 30, 2012 11:23 PM Robert Haas
[mailto:robertmh...@gmail.com] wrote:
On Fri, Aug 10, 2012 at 1:25 AM, Amit Kapila wrote:
>>> I think the property that recovery only needs to worry about each
>>> block individually is one that we want to preserve. Supporting this
>>> optimizatin
On 08/30/2012 07:23 PM, Bruce Momjian wrote:
> On Thu, Jul 12, 2012 at 06:01:00PM -0700, Joe Conway wrote:
>> I'll take a look at the latter option sometime in the next few weeks and
>> submit for the next commitfest.
>
> Any news on this?
Not yet -- OBE. I'll try to set aside some time on the lo
On Thu, Jul 12, 2012 at 06:01:00PM -0700, Joe Conway wrote:
> On 07/12/2012 02:53 PM, Tom Lane wrote:
> > Peter Eisentraut writes:
> >> As long as we're spending time on this, I'd propose getting rid of
> >> lanplistrusted, at least for access checking. Instead, just don't
> >> install USAGE priv
psql has supported older servers for a great while now, so this sort of
things seems pretty useless now:
psql (9.2rc1, server 9.1.4)
WARNING: psql version 9.2, server version 9.1.
Some psql features might not work
I think it should be reduced to warning when connected to a newer
server.
--
S
On Thu, Jul 12, 2012 at 12:35:26AM +0300, Peter Eisentraut wrote:
> This might be useful for some people. Here is an emacs configuration
> for perl-mode that is compatible with the new perltidy settings. Note
> that the default perl-mode settings produce indentation that will be
> completely shre
On Thu, 2012-08-30 at 17:36 -0400, Tom Lane wrote:
> Joe Abbate writes:
> > As an aside, I installed jade (on Debian) and tried to make world but
> > got several errors, starting with the following:
>
> > jade -wall -wno-unused-param -wno-empty -wfully-tagged -D . -D . -d
> > stylesheet.dsl -t
On Thu, Aug 30, 2012 at 7:47 PM, Jaime Casanova wrote:
> On Thu, Aug 30, 2012 at 4:58 PM, Joe Abbate wrote:
>> On 30/08/12 17:36, Tom Lane wrote:
>>> FWIW, that suggests that this version of jade is too old. I'm not sure
>>> that jade per se (as opposed to the successor project openjade) can be
On Thu, Aug 30, 2012 at 4:58 PM, Joe Abbate wrote:
> On 30/08/12 17:36, Tom Lane wrote:
>> FWIW, that suggests that this version of jade is too old. I'm not sure
>> that jade per se (as opposed to the successor project openjade) can be
>> used to build our docs at all --- you should check whether
On Thu, Aug 30, 2012 at 2:05 PM, Jeff Janes wrote:
> On Thu, Aug 30, 2012 at 1:18 PM, Joe Abbate wrote:
>>
>>gmake world
>>
>> Unfortunately, that failed because the doc build requires jade. I
>> managed to build contrib separately, but wanted to point out that jade
>> is not mentioned in th
On Thu, Aug 30, 2012 at 3:48 PM, Tomas Vondra wrote:
> That sounds like a pretty trivial patch. I've been thinking about yet
> another option - histograms (regular or with exponential bins).
I thought about that, too, but I think high-outliers is a lot more
useful. At least for the kinds of thin
On Thu, Aug 30, 2012 at 4:51 PM, Tom Lane wrote:
> Bruce Momjian writes:
>> On Thu, May 31, 2012 at 09:20:43AM +0900, Tatsuo Ishii wrote:
Ok, I modified the part of pg_dump where tremendous number of LOCK
TABLE are issued. I replace them with single LOCK TABLE with multiple
tables.
On Thu, Aug 30, 2012 at 3:25 PM, Tomas Vondra wrote:
> On 30 Srpen 2012, 18:02, Robert Haas wrote:
>> On Fri, Aug 24, 2012 at 5:25 PM, Tomas Vondra wrote:
>>> This patch is a bit less polished (and more complex) than the other
>>> pgbench patch I've sent a while back, and I'm not sure how to hand
On Thu, Aug 30, 2012 at 4:48 PM, Tom Lane wrote:
> Robert Haas writes:
>> On Thu, Aug 30, 2012 at 4:15 PM, Tom Lane wrote:
>>> Yeah, the idea of replacing sum_grow with a boolean just occurred to me
>>> too. As is, I think the code is making some less-than-portable
>>> assumptions about what wi
On 30 Srpen 2012, 23:44, Robert Haas wrote:
> On Thu, Aug 30, 2012 at 3:48 PM, Tomas Vondra wrote:
>> That sounds like a pretty trivial patch. I've been thinking about yet
>> another option - histograms (regular or with exponential bins).
>
> I thought about that, too, but I think high-outliers is
On 30/08/12 17:36, Tom Lane wrote:
> FWIW, that suggests that this version of jade is too old. I'm not sure
> that jade per se (as opposed to the successor project openjade) can be
> used to build our docs at all --- you should check whether this is
> openjade, or really the original project.
It
On Sun, Jun 17, 2012 at 11:45:54PM +0800, Magnus Hagander wrote:
> On Sun, Jun 17, 2012 at 11:42 PM, Tom Lane wrote:
> > Magnus Hagander writes:
> >> Is there a reason why we don't have a parameter on the client
> >> mirroring ssl_ciphers?
> >
> > Dunno, do we need one? I am not sure what the ci
On 30 Srpen 2012, 23:47, Robert Haas wrote:
> On Thu, Aug 30, 2012 at 3:25 PM, Tomas Vondra wrote:
>> On 30 Srpen 2012, 18:02, Robert Haas wrote:
>>> On Fri, Aug 24, 2012 at 5:25 PM, Tomas Vondra wrote:
This patch is a bit less polished (and more complex) than the other
pgbench patch I'
On Thu, Aug 30, 2012 at 3:17 PM, Tomas Vondra wrote:
> On 30 Srpen 2012, 17:53, Robert Haas wrote:
>> On Fri, Aug 24, 2012 at 6:36 PM, Tomas Vondra wrote:
>>> attached is a patch that improves performance when dropping multiple
>>> tables within a transaction. Instead of scanning the shared buffe
On Sun, Jun 17, 2012 at 12:00:20AM -0400, nik9...@gmail.com wrote:
> I've always used -1-f - < file.sql. It is confusing that -1 doesn't warn you
> when it wont work though.
This will be fixed in 9.3 with this commit:
commit be690e291d59e8d0c9f4df59abe09f1ff6cc0da9
Author: Rober
Hello Jeff,
On 30/08/12 17:05, Jeff Janes wrote:
> I think is probably because you don't have "DocBook DTD" or some of
> the other prerequisites listed in the URL I gave above.
Indeed. I was able to build world after invoking the apt-get line in
J.2.3 on that page. The only adjustment I had to
Joe Abbate writes:
> As an aside, I installed jade (on Debian) and tried to make world but
> got several errors, starting with the following:
> jade -wall -wno-unused-param -wno-empty -wfully-tagged -D . -D . -d
> stylesheet.dsl -t sgml -i output-html -V html-index postgres.sgml
> jade:E: unkno
Excerpts from Joe Abbate's message of jue ago 30 16:18:05 -0400 2012:
> Hello hackers,
>
> In order to test 9.2rc1, I had to build contrib (because Pyrseas uses
> some of those modules). The build instructions
> (http://www.postgresql.org/docs/9.2/static/install-procedure.html )
> state the way t
On Thu, Aug 30, 2012 at 04:51:02PM -0400, Tom Lane wrote:
> Bruce Momjian writes:
> > On Wed, May 30, 2012 at 06:10:12PM -0400, Tom Lane wrote:
> >> However, as pointed out by Patric, if you dump and restore an old
> >> timestamptz value in one of these zones, it will fail to restore because
> >>
Bruce Momjian writes:
> On Thu, Aug 30, 2012 at 05:01:39PM -0400, Bruce Momjian wrote:
>> Oh, got -/_ mixed up. Fixed with attached applied patch.
> Oops, that text is talking about Python's configure, so I put the text
> back. Seemed we had _no_ mention of our own --enable-shared.
Yeah, I just
On Thu, Aug 30, 2012 at 04:50:22PM -0400, Tom Lane wrote:
> Bruce Momjian writes:
> > On Mon, May 28, 2012 at 02:27:15AM +0300, Peter Eisentraut wrote:
> >> We should just remove it now.
>
> > --disable-shared removed, with the attached, applied patch.
>
> No documentation changes?
I couldn't f
Bruce Momjian writes:
> On Thu, May 31, 2012 at 09:20:43AM +0900, Tatsuo Ishii wrote:
>>> Ok, I modified the part of pg_dump where tremendous number of LOCK
>>> TABLE are issued. I replace them with single LOCK TABLE with multiple
>>> tables. With 100k tables LOCK statements took 13 minutes in tot
On Thu, Aug 30, 2012 at 05:01:39PM -0400, Bruce Momjian wrote:
> On Thu, Aug 30, 2012 at 04:57:30PM -0400, Tom Lane wrote:
> > Bruce Momjian writes:
> > > On Thu, Aug 30, 2012 at 04:50:22PM -0400, Tom Lane wrote:
> > >> No documentation changes?
> >
> > > I couldn't find any place we document it.
On Thu, Aug 30, 2012 at 1:18 PM, Joe Abbate wrote:
> Hello hackers,
>
> In order to test 9.2rc1, I had to build contrib (because Pyrseas uses
> some of those modules). The build instructions
> (http://www.postgresql.org/docs/9.2/static/install-procedure.html )
> state the way to build everything
On Thu, Aug 30, 2012 at 04:51:56PM -0400, Tom Lane wrote:
> Bruce Momjian writes:
> > On Thu, May 31, 2012 at 09:20:43AM +0900, Tatsuo Ishii wrote:
> >>> Ok, I modified the part of pg_dump where tremendous number of LOCK
> >>> TABLE are issued. I replace them with single LOCK TABLE with multiple
>
Excerpts from Alvaro Herrera's message of mié ago 29 15:13:11 -0400 2012:
> Excerpts from Andres Freund's message of mié ago 29 12:10:17 -0400 2012:
>
> > > OK, scratch that thought then. So we seem to be down to choosing a new
> > > name for what we're going to take out of htup.h. If you don't
On Thu, Aug 30, 2012 at 04:57:30PM -0400, Tom Lane wrote:
> Bruce Momjian writes:
> > On Thu, Aug 30, 2012 at 04:50:22PM -0400, Tom Lane wrote:
> >> No documentation changes?
>
> > I couldn't find any place we document it. I did:
> > grep _shared *.sgml
> > and no hits were returned. Should
Bruce Momjian writes:
> On Thu, Aug 30, 2012 at 04:50:22PM -0400, Tom Lane wrote:
>> No documentation changes?
> I couldn't find any place we document it. I did:
> grep _shared *.sgml
> and no hits were returned. Should I search for something else?
It's --enable-shared, not --enable_shar
Bruce Momjian writes:
> On Wed, May 30, 2012 at 06:10:12PM -0400, Tom Lane wrote:
>> However, as pointed out by Patric, if you dump and restore an old
>> timestamptz value in one of these zones, it will fail to restore because
>> of the sanity check. I think therefore that we'd better enlarge the
Bruce Momjian writes:
> On Mon, May 28, 2012 at 02:27:15AM +0300, Peter Eisentraut wrote:
>> We should just remove it now.
> --disable-shared removed, with the attached, applied patch.
No documentation changes?
regards, tom lane
--
Sent via pgsql-hackers mailing list
Robert Haas writes:
> On Thu, Aug 30, 2012 at 4:15 PM, Tom Lane wrote:
>> Yeah, the idea of replacing sum_grow with a boolean just occurred to me
>> too. As is, I think the code is making some less-than-portable
>> assumptions about what will happen if sum_grow overflows; which can
>> definitely
On Thu, May 31, 2012 at 09:20:43AM +0900, Tatsuo Ishii wrote:
> >> Yeah, Jeff's experiments indicated that the remaining bottleneck is lock
> >> management in the server. What I fixed so far on the pg_dump side
> >> should be enough to let partial dumps run at reasonable speed even if
> >> the who
On Wed, May 30, 2012 at 06:10:12PM -0400, Tom Lane wrote:
> Currently, our datetime input code thinks that any UTC offset of more
> than 14:59:59 either way from Greenwich must be a mistake. However,
> after seeing Patric Bechtel's recent bug report, I went trolling in the
> Olson timezone files t
On Tue, May 29, 2012 at 03:54:59PM -0700, Mark Dilger wrote:
> I was imagining that this would be a trap for linux developers
> who saw nothing wrong with their code until it made it to the
> build/test farm. That's pretty far down the development
> process. Of course, it is also a trap in the ot
On Mon, May 28, 2012 at 02:27:15AM +0300, Peter Eisentraut wrote:
> On lör, 2012-05-26 at 15:53 -0400, Tom Lane wrote:
> > 2. Seeing that this is the first complaint since 9.0, should we decide
> > that --disable-shared is no longer worth supporting? Seems like we
> > should either make this case
On 30 August 2012 20:28, Robert Haas wrote:
> On Sat, Jul 28, 2012 at 4:09 PM, Jeff Janes wrote:
>> But it might be better yet to make ordinary index scans benefit from
>> effective_io_concurrency, but even if/when that gets done it would
>> probably still be worthwhile to make the planner unders
On Thu, Aug 30, 2012 at 4:15 PM, Tom Lane wrote:
> Robert Haas writes:
>> I noticed all that, but didn't feel like putting in the effort to make
>> it better. I would have been happy to have someone else pick up the
>> patch, but as it had been languishing I thought it would be better to
>> get
Hello hackers,
In order to test 9.2rc1, I had to build contrib (because Pyrseas uses
some of those modules). The build instructions
(http://www.postgresql.org/docs/9.2/static/install-procedure.html )
state the way to build everything (contrib + docs, etc.) is
gmake world
Unfortunately, that
Robert Haas writes:
> I noticed all that, but didn't feel like putting in the effort to make
> it better. I would have been happy to have someone else pick up the
> patch, but as it had been languishing I thought it would be better to
> get it committed more or less as it was than to wait for som
On Thu, Aug 30, 2012 at 3:27 PM, Tom Lane wrote:
> Robert Haas writes:
>> On Thu, Aug 30, 2012 at 1:27 PM, Tom Lane wrote:
>>> Should we backpatch that?
>
>> Arguably, yes. Does the patch look sane to you?
>
> I was afraid you'd ask that.
>
> [ studies code for awhile ... ]
>
> I think this fix
On 30 Srpen 2012, 17:46, Robert Haas wrote:
> On Sun, Aug 26, 2012 at 1:04 PM, Tomas Vondra wrote:
>> Attached is an improved patch, with a call to rand() replaced with
>> getrand().
>>
>> I was thinking about the counter but I'm not really sure how to handle
>> cases like "39%" - I'm not sure a p
On Sat, Jul 28, 2012 at 4:09 PM, Jeff Janes wrote:
> From my attempted reading of the thread "posix_fadvise v22", it seems
> like modification of the planner was never discussed, rather than
> being discussed and rejected. So, is there a reason not to make the
> planner take account of effective_
Robert Haas writes:
> On Thu, Aug 30, 2012 at 2:18 PM, Pavel Stehule
> wrote:
>> 2012/8/30 Robert Haas :
>>> On Tue, Aug 14, 2012 at 3:46 AM, Pavel Stehule
>>> wrote:
patch that implements "shared" client/server session variables
>>> I don't really see what we can do with this that we ca
On Mon, Aug 13, 2012 at 6:19 PM, Jeff Janes wrote:
> On Sat, Jun 9, 2012 at 5:43 AM, Tom Lane wrote:
>> Simon Riggs writes:
>>> So now the standard for my patches is that I must consider what will
>>> happen if the xlog is deleted?
>>
>> When you're messing around with something that affects dat
On Thu, Jul 26, 2012 at 11:39 PM, wrote:
> Here is my task situation:
>
> I have a TupleTableSlot, with its own TupleDesc. Now I want to extract
> several attributes to form a new TupleTableSlot, how can I define my own
> TupleDesc and the ProjectionInfo?
You might get more helpful advice if you
Robert Haas writes:
> On Thu, Aug 30, 2012 at 1:27 PM, Tom Lane wrote:
>> Should we backpatch that?
> Arguably, yes. Does the patch look sane to you?
I was afraid you'd ask that.
[ studies code for awhile ... ]
I think this fixes the bug, but the function could really do with slightly
more i
On 30 Srpen 2012, 18:02, Robert Haas wrote:
> On Fri, Aug 24, 2012 at 5:25 PM, Tomas Vondra wrote:
>> This patch is a bit less polished (and more complex) than the other
>> pgbench patch I've sent a while back, and I'm not sure how to handle the
>> Windows branch. That needs to be fixed during the
On Mon, Aug 13, 2012 at 3:35 PM, Kohei KaiGai wrote:
> The attached patch is a refreshed version of ALTER command
> reworks towards the latest tree. Here is few big changes except
> for code integration of the code to rename event triggers.
This seems to have bit-rotted a bit. Please rebase.
>
On 30 Srpen 2012, 17:53, Robert Haas wrote:
> On Fri, Aug 24, 2012 at 6:36 PM, Tomas Vondra wrote:
>> attached is a patch that improves performance when dropping multiple
>> tables within a transaction. Instead of scanning the shared buffers for
>> each table separately, the patch removes this and
On Sun, Aug 12, 2012 at 5:14 PM, Dean Rasheed wrote:
> None of this new code kicks in for non-security barrier views, so the
> kinds of plans I posted upthread remain unchanged in that case. But
> now a significant fraction of the patch is code added to handle
> security barrier views. Of course w
On Thu, Aug 30, 2012 at 2:18 PM, Pavel Stehule wrote:
> 2012/8/30 Robert Haas :
>> On Tue, Aug 14, 2012 at 3:46 AM, Pavel Stehule
>> wrote:
>>> patch that implements "shared" client/server session variables
>>
>> I don't really see what we can do with this that we can't do without this.
>
> a mo
On Thu, Aug 9, 2012 at 1:11 PM, Kevin Grittner
wrote:
> I didn't figure it was; my emphasis was because this has been raised
> before and nothing happened for want of a consensus on what
> particular wording should be used, so users were left with no
> guidance. I don't want this to continue to b
Joe Abbate writes:
> Yes, I suspected that an OID was stored. What I'd still quibble with is
> the use of the ambiguous regproc in pg_operator (also pg_type) and the
> still-ambiguous schema-qualified proc name. I guess it's not feasible
> (at least, short term), but it'd be preferable to store
Hello Tom,
On 30/08/12 13:23, Tom Lane wrote:
> Joe Abbate writes:
>> Hmmm ... Well, I'm just doing the same thing as pg_dump, which in 9.2rc1
>> still outputs the same as before, namely:
>
> Well, evidently you're *not* doing the same thing as pg_dump.
I meant that the Pyrseas dbtoyaml's outpu
2012/8/30 Robert Haas :
> On Tue, Aug 14, 2012 at 3:46 AM, Pavel Stehule
> wrote:
>> patch that implements "shared" client/server session variables
>
> I don't really see what we can do with this that we can't do without this.
a motivation for this patch was discussion about parametrised DO
stat
On Tue, Aug 14, 2012 at 3:46 AM, Pavel Stehule wrote:
> patch that implements "shared" client/server session variables
I don't really see what we can do with this that we can't do without this.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
--
Sent
On Fri, Aug 10, 2012 at 1:25 AM, Amit Kapila wrote:
>> I think the property that recovery only needs to worry about each
>> block individually is one that we want to preserve. Supporting this
>> optimizating only when full_page_writes=off seems ugly,
>
> I think recovery needs to worry about mult
On Thu, Aug 30, 2012 at 9:11 PM, Robert Haas wrote:
> On Mon, Aug 20, 2012 at 12:57 PM, Alexander Korotkov
> wrote:
> > I found that code of gistchoose doesn't follow it's logic. Idea of
> > gistchoose is that first column penalty is more important than penalty of
> > second column. If we meet s
On Thu, Aug 30, 2012 at 12:14 PM, Andres Freund wrote:
>> An alternative thing that might be worth considering before you go all
>> in on this is whether the xlogdump functionality shouldn't just be part
>> of the regular server executable, ie you'd call it with
>>
>> postgres --xlogdump
>>
Andres Freund writes:
> On Thursday, August 30, 2012 06:50:13 PM Matthias wrote:
>> According to the debugger num_hist = 1, so it divides by zero.
> Its curious though that the SIGFPE isn't properly cought though. That would
> only lead to a different error, but ...
Not all platforms think an i
On Tue, Aug 14, 2012 at 5:46 PM, Tom Lane wrote:
> Noah Misch writes:
>> On Tue, Aug 14, 2012 at 08:40:06AM -0400, Robert Haas wrote:
>>> On Tue, Aug 14, 2012 at 6:50 AM, Greg Stark wrote:
It is possible to check if the signal was synchronous or was sent from
an external process. You c
On Thu, Aug 30, 2012 at 1:27 PM, Tom Lane wrote:
> Robert Haas writes:
>> On Mon, Aug 20, 2012 at 12:57 PM, Alexander Korotkov
>> wrote:
>>> I found that code of gistchoose doesn't follow it's logic. Idea of
>>> gistchoose is that first column penalty is more important than penalty of
>>> second
On Mon, Aug 13, 2012 at 11:59 PM, Peter Eisentraut wrote:
> On Tue, 2012-07-24 at 16:14 -0400, Robert Haas wrote:
>> On Tue, Jul 24, 2012 at 4:09 PM, Tom Lane wrote:
>> > Robert Haas writes:
>> >> Yeah, you're right. So you do get the table name. But you don't get
>> >> the cause, which is wha
On 30.08.2012 19:50, Matthias wrote:
It crashes in rangetypes_typeanalyze.c at line 186:
delta = (non_empty_cnt - 1) / (num_hist - 1);
According to the debugger num_hist = 1, so it divides by zero. I guess
this is due to the new statistics collection for range types?
Yep. Fixed, thanks f
Robert Haas writes:
> On Mon, Aug 20, 2012 at 12:57 PM, Alexander Korotkov
> wrote:
>> I found that code of gistchoose doesn't follow it's logic. Idea of
>> gistchoose is that first column penalty is more important than penalty of
>> second column. If we meet same penalty values of first column t
Joe Abbate writes:
> On 30/08/12 12:27, Tom Lane wrote:
>> The reason for the difference is that in 9.2 there's more than one
>> pg_catalog.upper():
> Hmmm ... Well, I'm just doing the same thing as pg_dump, which in 9.2rc1
> still outputs the same as before, namely:
Well, evidently you're *not*
On Thu, Aug 16, 2012 at 2:11 AM, Etsuro Fujita
wrote:
> Agreed. I'd like to withdraw the patch sent in the earlier post, and propose
> to
> update the documentation in the COPY reference page. Please find attached a
> patch.
I think this is a good idea, but I didn't like the exact wording you
On Mon, Aug 20, 2012 at 12:57 PM, Alexander Korotkov
wrote:
> I found that code of gistchoose doesn't follow it's logic. Idea of
> gistchoose is that first column penalty is more important than penalty of
> second column. If we meet same penalty values of first column then we choose
> minimal pena
Hello Tom,
On 30/08/12 12:27, Tom Lane wrote:
> The reason for the difference is that in 9.2 there's more than one
> pg_catalog.upper():
>
> regression=# \df upper
> List of functions
>Schema | Name | Result data type | Argument data types | Type
>
2012/8/30 Albe Laurenz :
> Matthias wrote:
>> when running VACUUM ANALYZE on my database built on win32-x86 from
>> yesterday's git checkout I always get this at some point during VACUUM
>> ANALYZE:
>>
>> LOG: server process (PID 5880) was terminated by exception 0xC094
>> DETAIL: Failed proc
On Thursday, August 30, 2012 06:50:13 PM Matthias wrote:
> 2012/8/30 Albe Laurenz :
> > Matthias wrote:
> >> when running VACUUM ANALYZE on my database built on win32-x86 from
> >> yesterday's git checkout I always get this at some point during VACUUM
> >> ANALYZE:
> >>
> >> LOG: server process (
Joe Abbate writes:
> Hello hackers,
> I've been testing Pyrseas against 9.2rc1. A test that does a CREATE
> OPERATOR is giving a small difference. Specifically, the test issues
> the statement:
> CREATE OPERATOR + (PROCEDURE = upper, RIGHTARG = text);
> Pyrseas then queries the pg_operator cat
On Thursday, August 30, 2012 06:09:43 PM Andres Freund wrote:
> On Thursday, August 30, 2012 06:06:59 PM Robert Haas wrote:
> > On Thu, Aug 30, 2012 at 10:36 AM, Miroslav Šimulčík
> >
> > wrote:
> > > is there any way to check if row have already been modified by the
> > > current transaction? I
Robert Haas writes:
> On Thu, Aug 30, 2012 at 10:36 AM, Miroslav Å imulÄÃk
> wrote:
>> is there any way to check if row have already been modified by the current
>> transaction? I tried condition txid_current() = xmin, but there is problem
>> with the savepoints. After every savepoint rows are
On Wednesday, August 29, 2012 10:06:16 PM Tom Lane wrote:
> Alvaro Herrera writes:
> > I looked at Andres' patch and the general idea is rather horrible: it
> > links all backend files into the output executable. This is so that the
> > *_desc functions can be used from their respective object fi
On Thursday, August 30, 2012 06:06:59 PM Robert Haas wrote:
> On Thu, Aug 30, 2012 at 10:36 AM, Miroslav Šimulčík
>
> wrote:
> > is there any way to check if row have already been modified by the
> > current transaction? I tried condition txid_current() = xmin, but there
> > is problem with the s
On Thu, Aug 30, 2012 at 10:36 AM, Miroslav Šimulčík
wrote:
> is there any way to check if row have already been modified by the current
> transaction? I tried condition txid_current() = xmin, but there is problem
> with the savepoints. After every savepoint rows are getting higher xmin
> values, b
On Fri, Aug 24, 2012 at 5:25 PM, Tomas Vondra wrote:
> This patch is a bit less polished (and more complex) than the other
> pgbench patch I've sent a while back, and I'm not sure how to handle the
> Windows branch. That needs to be fixed during the commit fest.
What's the problem with the Window
On Fri, Aug 24, 2012 at 6:36 PM, Tomas Vondra wrote:
> attached is a patch that improves performance when dropping multiple
> tables within a transaction. Instead of scanning the shared buffers for
> each table separately, the patch removes this and evicts all the tables
> in a single pass through
On Sun, Aug 26, 2012 at 1:04 PM, Tomas Vondra wrote:
> Attached is an improved patch, with a call to rand() replaced with
> getrand().
>
> I was thinking about the counter but I'm not really sure how to handle
> cases like "39%" - I'm not sure a plain (counter % 100 < 37) is not a
> good sampling,
On Mon, Aug 27, 2012 at 8:31 AM, Dirk Lutzebäck
wrote:
> we have issues with compound words in tsearch2 using the german (ispell)
> dictionary. This has been discussed before but there is no real solution
> using the recommended german dictionary at
> http://www.sai.msu.su/~megera/postgres/gist/ts
On Mon, Aug 27, 2012 at 7:52 AM, Dimitri Fontaine
wrote:
> I've been reviewing your changes and here's a very small patch with some
> details I would have spelled out differently. See what you think, I
> mostly needed to edit some code to get back in shape :)
I guess I don't particularly like eit
Hello hackers,
I've been testing Pyrseas against 9.2rc1. A test that does a CREATE
OPERATOR is giving a small difference. Specifically, the test issues
the statement:
CREATE OPERATOR + (PROCEDURE = upper, RIGHTARG = text);
Pyrseas then queries the pg_operator catalog to map the procedure for
o
On Mon, Aug 27, 2012 at 11:21 PM, Jeff Davis wrote:
> I noticed a couple comments that look wrong to me. Patch attached.
Thanks, committed. But I updated the parenthesized comment in the
first fix instead of removing it. Let me know if you see an issue
with that.
--
Robert Haas
EnterpriseDB:
Hi,
is there any way to check if row have already been modified by the current
transaction? I tried condition txid_current() = xmin, but there is problem
with the savepoints. After every savepoint rows are getting higher xmin
values, but txid_current() remains the same.
Regards,
Miroslav Simulcik
Daniel Farina writes:
> but just today we promoted another system via streaming replication to
> pick up the planner fix in 9.1.5 (did you know: that planner bug seems
> to make GIN FTS indexes un-used in non-exotic cases, and one goes to
> seqscan?), and then a 40MB GIN index bloated to two gigs
On Thu, Aug 30, 2012 at 2:00 AM, Matthias wrote:
> Hey,
>
> I tried out the new CHECK NO INHERIT feature for inherited tables.
> There seems to be an opportunity to generate slightly better query
> plans sometimes. E.g. when I do
>
> SELECT * FROM base WHERE partition_id = 3
>
> and there exists o
On Wed, Aug 29, 2012 at 10:25 PM, Peter Geoghegan wrote:
> On 19 February 2012 05:24, Robert Haas wrote:
>> I have attached tps scatterplots. The obvious conclusion appears to
>> be that, with only 16MB of wal_buffers, the buffer "wraps around" with
>> some regularity: we can't insert more WAL b
On Tue, May 22, 2012 at 08:29:48AM -0400, Robert Haas wrote:
> On Tue, May 22, 2012 at 1:50 AM, Tom Lane wrote:
> > Currently, the planner keeps paths that appear to win on the grounds of
> > either cheapest startup cost or cheapest total cost. It suddenly struck
> > me that in many simple cases
On Wed, Aug 29, 2012 at 6:39 PM, Tom Lane wrote:
> Well, I see your point about LPAD(), but the problem is how to tell
> the difference between a harmless cast omission and an actual mistake
> that the user will be very grateful if we point out. If we allow
> implicit casts to text in the general
1 - 100 of 109 matches
Mail list logo
