Hi,
I just noticed plpgsql evaluates all AND'ed conditions even if the first
one fails. Example:
elsif TG_OP = ''UPDATE'' and old.type_reponse = ''abandon''
This will break stuff if the trigger is used on INSERT as
"old.type_reponse" will be substituted and return an error.
Shouldn't
On Mon, May 27, 2002 at 05:12:33PM -0400, Bruce Momjian wrote:
> Last I talked to Darren, the replication code was modified to merge into
> our 7.2 tree. There are still pieces missing so it will not be
> functional when applied. It is remotely possible there could be
> master-slave in 7.3, but
Tom Lane wrote:
> Teodor Sigaev <[EMAIL PROTECTED]> writes:
>
>>>Yeah, but the update case is inserting more entries into the index.
>>>I'm wondering if that causes the index scan's state to get corrupted
>>>so that it misses scanning some entries.
>>>
>
>>Thank you, Tom. You give me a directi
uff, I am ashamed
Tracing the problem I found out that the invalid startup packets were not
triggered by the webserver either... looking more precisely I found out
that the messages appeared regularly every 180 seconds in /var/log/messages
This led me to the thought, that this has got to be
On Mon, 27 May 2002 13:48:43 -0400, Tom Lane <[EMAIL PROTECTED]>
wrote:
>Manfred Koizar <[EMAIL PROTECTED]> writes:
>> On leaf pages order index tuples by heap item pointer, if otherwise
>> equal. [blah, blah, ...]
>
>Doesn't help, I fear. Finding your place again is only one part
>of the proble
On Mon, May 27, 2002 at 11:40:20AM -0400, Michael Meskes wrote:
> could anyone please enlighten me about the status of replication? I do
> expect lots of questions about this, and I'm not really sure if I can
> promise it for 7.3. :-)
8.0 ;-) (?)
I add the other quesion: how is current status
> FOR row IN select_query LOOP
> statements
> RETURN NEXT row;
> END LOOP;
Informix has
RETURN x1, x2, x3 WITH RESUME;
This seems reasonable to me. PostgreSQL could also allow
return x with resume, where x is already a composite type.
Andreas
---(end
Hi Bruce,
Thank you for your reply. It makes a lot of sense!
However I don't really understand why we can't control the NUMBER of
files.
Are the 8 files I see a maximum usage when I reloaded the databases on the
ne system or is it some sort of "plugged in value"?
Thank you for your explanation.
Actually, at least in some cases, PG does short-circuit logic:
create function seeme() returns bool as '
begin
raise notice ''seeme'';
return true;
end'
language plpgsql;
joel@joel=# select false and seeme();
?column?
--
f
(1 row)
joel@joel=# select true and seeme();
NOTIC
Teodor Sigaev <[EMAIL PROTECTED]> writes:
>> Hmm, is this patch really correct? Removing the gistadjscans() call
>> from gistSplit seems wrong to me --- won't that miss reporting splits
>> on leaf pages? Or does this not matter for some reason?
> gistadjscans() is moving to gistlayerinsert. gis
Louis-David Mitterrand <[EMAIL PROTECTED]> writes:
> I just noticed plpgsql evaluates all AND'ed conditions even if the first
> one fails. Example:
> elsif TG_OP = ''UPDATE'' and old.type_reponse = ''abandon''
> This will break stuff if the trigger is used on INSERT as
> "old.type_reponse"
Joe Conway <[EMAIL PROTECTED]> writes:
> What if we also had something like:
> FUNC_BUILD_TUPLE(values, funcctx);
> which returns a tuple for the less experienced folks
Sure, as long as it's not getting in the way when you don't want it.
For that matter the FUNC stuff shouldn't get in the way
On Tue, May 28, 2002 at 09:53:53AM +0200, Karel Zak wrote:
> On Mon, May 27, 2002 at 11:40:20AM -0400, Michael Meskes wrote:
>
> > could anyone please enlighten me about the status of replication? I do
> > expect lots of questions about this, and I'm not really sure if I can
> > promise it for 7.
Louis-David Mitterrand writes:
> Shouldn't plpgsql shortcut AND conditions when a previous one fails, as
> perl does?
Shouldn't perl evaluate all operands unconditionally, like plpgsql does?
Seriously, if you want to change this you have to complain to the SQL
standards committee.
--
Peter Ei
On Tue, 28 May 2002, Michael Meskes wrote:
>
> This is about pgreplication I think. Is the the replication project of
> choice for pgsql? IIRC there quite some projects for this topic:
>
> PostgreSQL replicator
> Rserver
> Usogres
> dbbalancer
There's also DBMirror which I submitted to the co
This feature does not affect the original version of poly_overlap as only a bounding
box test is preformed. I modified poly_overlap in an attempt to improve the
preciseness of poly_overlap. The function works when the column is not indexed or
when the column is indexed using rtree_gist fro
With 7.1.3, large indexes with null values allowed in one or more of the
columns would cause crashes. (I have definitely seen this happen).
Here is a project that mentions repairs:
http://postgis.refractions.net/news/index.php?file=20020425.data
Have repairs been effected in 7.2? Are they dela
"Kenneth Chan" <[EMAIL PROTECTED]> writes:
> ... Turned out that npts of the
> polygon retrieved from the table is 0 (the other polygon is a constant
> and its attributes are correct). I suspect the feature might
> affect other functions that uses polygons->npts like poly_contain.
> Would anyon
"Dann Corbit" <[EMAIL PROTECTED]> writes:
> With 7.1.3, large indexes with null values allowed in one or more of the
> columns would cause crashes. (I have definitely seen this happen).
> Have repairs been effected in 7.2?
Submit a test case and we'll tell you ...
regard
> > > > ./configure --help
> > > > Isn't anywhere else yet.
> > > Not seeing it. Is this a 7.3 thing? What does it do?
> > Sorry, yes it is a 7.3 thing.
> What does --enable-interval-integers do? I don't want to bother writing up
> issues you've already taken care of.
Not implemented afaik. Or
The Perl build (PL/Perl and the Pg interface) now use the configured
compiler and flags and none of the MakeMaker stuff. (I've kept the
interfaces/perl5/Makefile.PL file in case someone wants to resurrect it
for a Win32 build, for instance.) Since doing Perl builds without
MakeMaker is poorly do
On Tue, 28 May 2002, Tom Lane wrote:
> Teodor Sigaev <[EMAIL PROTECTED]> writes:
> > Thank you, Tom. You give me a direction for looking. Attached patch fix
> > the problem with broken state. Please apply it for 7.2.2 and current cvs
>
> Patch applied to current and REL7_2 branch.
Is't time for
I'm trying to implement some code to recreate tables as we discussed
formerly. But it's not so easy... :-) My first blind alley is that
dropping a function which is occured in a CHECK constraint or
a DEFAULT constraint, I get "fmgr_info: function 12345678: cache
lookup failed" or "Function OID 123
On Mon, May 27, 2002 at 09:00:43PM -0400, Bruce Momjian wrote:
>
> TODO updated:
>
> Add BSD-licensed qsort() for 32-bit Solaris
I've received an email noting that someone else ran a test program
with the 64 bit library, and had just as bad performance as the 32
bit one. I haven't had a
Kovacs Zoltan <[EMAIL PROTECTED]> writes:
> I'm trying to implement some code to recreate tables as we discussed
> formerly. But it's not so easy... :-) My first blind alley is that
> dropping a function which is occured in a CHECK constraint or
> a DEFAULT constraint, I get "fmgr_info: function 1
Tom Lane wrote:
> Teodor Sigaev <[EMAIL PROTECTED]> writes:
>
>>>Hmm, is this patch really correct? Removing the gistadjscans() call
>>>from gistSplit seems wrong to me --- won't that miss reporting splits
>>>on leaf pages? Or does this not matter for some reason?
>>
>
>>gistadjscans() is mo
Teodor Sigaev <[EMAIL PROTECTED]> writes:
> Internally splits are doing before calling gistadjscans. All pages
> created by gistSplit will be inserted in the end of parent page.
> GiST's indexes aren't a concurrent there for one call of gistadjscans
> will be sufficiant.
Oh, I see. Thanks.
On Tue, 28 May 2002, Tom Lane wrote:
> Kovacs Zoltan <[EMAIL PROTECTED]> writes:
> > I'm trying to implement some code to recreate tables as we discussed
> > formerly. But it's not so easy... :-) My first blind alley is that
> > dropping a function which is occured in a CHECK constraint or
> > a
Teodor Sigaev <[EMAIL PROTECTED]> writes:
> Thank you, Tom. You give me a direction for looking. Attached patch fix
> the problem with broken state. Please apply it for 7.2.2 and current cvs
Patch applied to current and REL7_2 branch.
regards, tom lane
-
Teodor Sigaev <[EMAIL PROTECTED]> writes:
>> Attached patch fix a bug with creating index. Bug was reported by Chris
>> Hodgson <[EMAIL PROTECTED]>. Please, apply it for 7.2.2 and
>> current CVS.
Patch applied to both branches.
regards, tom lane
---
Oleg Bartunov <[EMAIL PROTECTED]> writes:
> Is't time for 7.2.2 ?
I think we had agreed start of June for 7.2.2.
regards, tom lane
---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Thomas,
> > > ./configure --help
> > > Isn't anywhere else yet.
> > Not seeing it. Is this a 7.3 thing? What does it do?
>
> Sorry, yes it is a 7.3 thing.
What does --enable-interval-integers do? I don't want to bother writing up
issues you've already taken care of.
--
-Josh Berkus
On Tue, 2002-05-28 at 21:52, Peter Eisentraut wrote:
> Louis-David Mitterrand writes:
>
> > Shouldn't plpgsql shortcut AND conditions when a previous one fails, as
> > perl does?
>
> Shouldn't perl evaluate all operands unconditionally, like plpgsql does?
>
> Seriously, if you want to change th
On Wed, 2002-05-29 at 02:36, Joel Burton wrote:
> > -Original Message-
> joel@joel=# select true and seeme();
> NOTICE: seeme
> ?column?
> --
> t
> (1 row)
>
>
> It certainly appears to be short circuiting for "select false and seeme()",
> for instance.
>
> It appears that th
Hannu Krosing <[EMAIL PROTECTED]> writes:
> How hard would it be to _not_ include nulls in indexes
> as they are not used anyway.
Seems to me that would be a step backwards.
What should someday happen is to make IS NULL an indexable operator.
The fact that we haven't got around to doing so is
Hannu Krosing <[EMAIL PROTECTED]> writes:
> Are these intricacies of SQL standardised anywhere ?
SQL92 section 3.3.4.4, "rule evaluation order" appears to sanction PG's
behavior. In particular note the part that says syntax rules and access
rules are "effectively applied at the same time" (ie, t
Neil Conway wrote:
> On Mon, 27 May 2002 21:00:43 -0400 (EDT)
> "Bruce Momjian" <[EMAIL PROTECTED]> wrote:
> > TODO updated:
> >
> > Add BSD-licensed qsort() for 32-bit Solaris
>
> Is this necessary? Didn't someone say that Sun had acknowledged the
> performance problem and were going to be
Andrew Sullivan wrote:
> On Mon, May 27, 2002 at 09:00:43PM -0400, Bruce Momjian wrote:
> >
> > TODO updated:
> >
> > Add BSD-licensed qsort() for 32-bit Solaris
>
> I've received an email noting that someone else ran a test program
> with the 64 bit library, and had just as bad performanc
Karel Zak wrote:
> On Mon, May 27, 2002 at 11:40:20AM -0400, Michael Meskes wrote:
>
> > could anyone please enlighten me about the status of replication? I do
> > expect lots of questions about this, and I'm not really sure if I can
> > promise it for 7.3. :-)
>
> 8.0 ;-) (?)
>
> I add the o
> -Original Message-
> From: Bruce Momjian [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, May 28, 2002 5:37 PM
> To: Andrew Sullivan
> Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Re: [HACKERS] [GENERAL] Re : Solaris Performance - 64 bit
> puzzle
>
>
> Andrew Sullivan wrote:
> > On M
Michael Meskes wrote:
> On Mon, May 27, 2002 at 05:12:33PM -0400, Bruce Momjian wrote:
> > Last I talked to Darren, the replication code was modified to merge into
> > our 7.2 tree. There are still pieces missing so it will not be
> > functional when applied. It is remotely possible there could
> ... Turned out that npts of the
> polygon retrieved from the table is 0 (the other polygon is a constant
> and its attributes are correct). I suspect the feature might
> affect other functions that uses polygons->npts like poly_contain.
> Would anyone happens to know the identity of the offe
8 is the maximum unless WAL files have to be created _while_ the
checkpoint is taking place.
Current CVS SGML has:
The number of 16MB segment files will always be at least
WAL_FILES + 1, and will normally not exceed
WAL_FILES + MAX(WAL_FILES,
CHECKPOINT_SEGMENTS) + 1.
The real driv
...
> rserver only does single-master, while most people want multi-master.
As you probably know, rserv is not limited to only a single instance of
a single master. Many replication problems can be described as a "single
source" problem (or should be described as such; moving to a fully
distribut
Agreed. It would be nice to see both a single-master and multi-master
server included in our main tree and a clear description of when to use
each. The confusion over the various replication solutions and their
strengths/weaknesses is a major problem.
I always felt a clearer README for rserv w
Tom Lane wrote:
> Neil Conway <[EMAIL PROTECTED]> writes:
> > Hmmm... Well, I'll take a look at it, but I'll probably just leave it
> > be -- since the optimization might actually return invalid results, it
> > doesn't seem like a very valuable thing to have, IMHO.
>
> Yeah, I never cared for the
46 matches
Mail list logo
