How does Index scan perform a scan for overlapping Index Cond ?
If I get a plan like this, what will actually be performed if EXPLAIN
shows this:
Sort (cost=12.90..12.91 rows=1 width=207)
Sort Key: log_actionseq
-> Index Scan using sl_log_1_idx2_hu, sl_log_1_idx2_hu,
sl_log_1_idx2_hu, sl
Tom Lane <[EMAIL PROTECTED]> wrote:
> #define MAX_TUPLES_PER_PAGE ((BLCKSZ - 1) /
> (MAXALIGN(offsetof(HeapTupleHeaderData, t_bits)) + sizeof(ItemIdData)) + 1)
> (I believe that both modules want a ceiling definition not a floor
> definition, ie round up any fraction. The -1 / +1 trick is of cou
Am Donnerstag, den 01.09.2005, 23:34 -0500 schrieb Jim C. Nasby:
> On Thu, Sep 01, 2005 at 11:18:25PM +0100, Simon Riggs wrote:
> As a side note, there's a typo in \?:
>
> \c[onnect] [DBNAME|- [USER]]
>
> Note the | where there should be a ].
No ;) The | stands for the alternative.
The whole
On Thu, Sep 01, 2005 at 11:34:37PM -0500, Jim C. Nasby wrote:
> As a side note, there's a typo in \?:
>
> \c[onnect] [DBNAME|- [USER]]
>
> Note the | where there should be a ].
Eh? Looks right to me; the | indicates an alternate, i.e., that
you can use either DBNAME or -. I often use - to co
On Thu, Sep 01, 2005 at 11:18:25PM +0100, Simon Riggs wrote:
> On Thu, 2005-09-01 at 14:53 -0400, Tom Lane wrote:
> > Darcy Buskermolen <[EMAIL PROTECTED]> writes:
> > > Yes I've seen the same frustration from users, I think the banner should
> > > say
> > > something along the lines of "psql cli
ITAGAKI Takahiro <[EMAIL PROTECTED]> writes:
> I found two different definitions of MAX_TUPLES_PER_PAGE.
> Which is reasonable? Or do they have another meaning?
Hmm, I think those were both my fault at different times :-(.
Right now I am thinking that they are both not quite right,
in particular i
On Thu, Sep 01, 2005 at 11:22:07PM -0400, Tom Lane wrote:
> "Jim C. Nasby" <[EMAIL PROTECTED]> writes:
> > On Thu, Sep 01, 2005 at 09:22:35AM -0400, Tom Lane wrote:
> >> This is a bad idea. The bgwriter isn't the place to be doing freezing,
>
> > So is this something that another daemon could han
"Jim C. Nasby" <[EMAIL PROTECTED]> writes:
> On Thu, Sep 01, 2005 at 09:22:35AM -0400, Tom Lane wrote:
>> This is a bad idea. The bgwriter isn't the place to be doing freezing,
> So is this something that another daemon could handle?
Possibly, but I'd be inclined to think of it as autovacuum's p
On Thu, Sep 01, 2005 at 09:22:35AM -0400, Tom Lane wrote:
> ITAGAKI Takahiro <[EMAIL PROTECTED]> writes:
> > I agree. I think an good position of freezer is on bgwriter.
> > My idea is:
> > 1. Just before bgwriter writes an dirty page in LRU order,
> > 2. Freeze tuples in the page and repair fr
On Thu, Sep 01, 2005 at 09:06:29PM -0500, Jim C. Nasby wrote:
> On Wed, Aug 31, 2005 at 04:08:10PM -0400, huaxin zhang wrote:
> > I am new to this hacker's job. What I was looking for was to record
> > the actual disk IO performed for arbituary query plan. I searched
> > in backend/executor but not
On Wed, Aug 31, 2005 at 04:08:10PM -0400, huaxin zhang wrote:
> I am new to this hacker's job. What I was looking for was to record
> the actual disk IO performed for arbituary query plan. I searched
> in backend/executor but not sure if that was the right place to
> add a tracer. would the /backe
Title: RE: PL/pgSQL: EXCEPTION NOSAVEPOINT
>if I could get into the TRY section of the PG_CATCH()/PG_TRY()
> construct without an intervening elog(ERROR) then I'd have a
> chance ...
Sorry, I meant "the CATCH section of the PG_TRY()/PG_CATCH()
construct."
Hi,
I found two different definitions of MAX_TUPLES_PER_PAGE.
Which is reasonable? Or do they have another meaning?
backend/commands/vacuumlazy.c
#define MAX_TUPLES_PER_PAGE ((int) (BLCKSZ / sizeof(HeapTupleHeaderData)))
backend/nodes/tidbitmap.c
#define MAX_TUPLES_PER_PAGE ((BLCKSZ - 1) /
MA
Title: RE: PL/pgSQL: EXCEPTION NOSAVEPOINT
> In general I don't think it even makes sense to think of making
> executor rollback non-transactional.
Agreed.
I would not want to rollback some statements and not others within a
transaction. I would like a complete rollback to happen, but onl
Why don't you just use EnterpriseDB?
Chris
That would defeat my goal of not rewriting all my Oracle code.
If I were fool enough to plan an attack on the main executor's exception
handling to try and disarm it of its subtransaction semantics, where
would I start? Where would I end? What would
Hi.
In a cluster, is there any way to use the main memory of the other nodes instead of the swap? If I have a query with many sub-queries and a lot of data, I can easily fill all the memory in a node. The point is: is there any way to continue using the main memory from other nodes in the same qu
Tom Lane wrote:
> Hannu Krosing <[EMAIL PROTECTED]> writes:
> > My wild guess is that deleting all index pointers for a removed index is
> > more-or-less the same cost as creating new ones for inserted/updated
> > page.
>
> Only if you are willing to make the removal process recalculate the
> inde
Bruce Momjian writes:
> I assume that for a vacuum that only hit pages indicated in the bitmap,
> it would still be necessary to do an index scan to remove the heap
> pointers in the index, right?
Given the current vacuum technology, yes. However, bearing in mind that
indexes should generally be
Matt Miller <[EMAIL PROTECTED]> writes:
> If I were fool enough to plan an attack on the main executor's exception
> handling to try and disarm it of its subtransaction semantics, where
> would I start? Where would I end? What would I do in between? Can New
> Orleans be rebuilt above sea level?
On Thu, 1 Sep 2005, Darcy Buskermolen wrote:
Yes I've seen the same frustration from users, I think the banner should say
something along the lines of "psql client version foo connected to server bar
version sfoo"
I second this, I think that something like "psl client version
8.03 connected
On Thu, 2005-09-01 at 18:28 -0400, Tom Lane wrote:
> Matt Miller <[EMAIL PROTECTED]> writes:
> > Basically I'd like my Pl/pgSQL code to be able to utilize the try/catch
> > paradigm of error handling without the overhead of subtransactions
>
> [Pl/pgSQL] can't even do 2+2 without
> calling the ma
On Thu, 1 Sep 2005, Tom Lane wrote:
Andrew Dunstan <[EMAIL PROTECTED]> writes:
Tom Lane wrote:
Sure you can. Make a SQL or PLPGSQL function that does the conversion
you want and then create a cast using it.
That probably won't help him much with "values(0)":
If I remember the context of
[ redirected to -hackers, where it's actually on topic ]
Matt Miller <[EMAIL PROTECTED]> writes:
> [redirected from -patches]
> On Wed, 2005-08-03 at 16:25 -0400, Tom Lane wrote:
>> This fundamentally breaks the entire backend. You do not have the
>> option to continue processing after elog(ERROR
On Thu, 2005-09-01 at 14:53 -0400, Tom Lane wrote:
> Darcy Buskermolen <[EMAIL PROTECTED]> writes:
> > Yes I've seen the same frustration from users, I think the banner should
> > say
> > something along the lines of "psql client version foo connected to server
> > bar
> > version sfoo"
>
> Th
On Thu, 2005-09-01 at 17:55 -0400, Tom Lane wrote:
> Allan Wang <[EMAIL PROTECTED]> writes:
> > Alright, I see why the checks are still needed. The unique index should
> > be on relname, conname right? Also looking into DROP CONSTRAINT's code,
> > it gives a notice about "multiple constraint names
Allan Wang <[EMAIL PROTECTED]> writes:
> (It would still be used for suppressing the not-exists error)
Right, which is why RemoveRelConstraints has to tell if it removed
anything. The API could be changed, but not to "returns void".
regards, tom lane
Allan Wang <[EMAIL PROTECTED]> writes:
> Alright, I see why the checks are still needed. The unique index should
> be on relname, conname right? Also looking into DROP CONSTRAINT's code,
> it gives a notice about "multiple constraint names dropped" when
> RemoveRelConstraints(rel, conname) returns
On Thu, 2005-09-01 at 17:16 -0400, Tom Lane wrote:
> Allan Wang <[EMAIL PROTECTED]> writes:
> > I've been looking through the code from CommentConstraint
> > and ATExecDropConstraint and they error out on duplicate constraint
> > names for a relation. However, ADD CONSTRAINT's code checks for
> > d
Martijn van Oosterhout writes:
> Do we have some platforms that don't have any multi-language support? I
> mean, we don't have a complete thread library but a wrapper around the
> ones used on the platform. Couldn't we make a similar wrapper that used
> glibc if it was available, windows native if
On Thu, Sep 01, 2005 at 02:31:15PM -0400, Tom Lane wrote:
> Patrick Welche <[EMAIL PROTECTED]> writes:
> > I think we still recommend using *new* pg_dump to dump *old* server when
> > upgrading. If one tries that with today's pg_dump (8.1beta1) against
> > a 8.1devel server of 6 May, i.e., predati
Andrew Dunstan <[EMAIL PROTECTED]> writes:
> Tom Lane wrote:
>> Sure you can. Make a SQL or PLPGSQL function that does the conversion
>> you want and then create a cast using it.
> That probably won't help him much with "values(0)":
If I remember the context of the original request, it wasn't ab
Tom Lane wrote:
"Marc G. Fournier" <[EMAIL PROTECTED]> writes:
On Mon, 29 Aug 2005, Tom Lane wrote:
No, because there's no built-in cast from smallint to bool.
'k, I just took a read through the "CREATE CAST" man page, and don't think
I can use that for this,
Sure
Allan Wang <[EMAIL PROTECTED]> writes:
> I've been looking through the code from CommentConstraint
> and ATExecDropConstraint and they error out on duplicate constraint
> names for a relation. However, ADD CONSTRAINT's code checks for
> duplicates and errors out, so would the stuff in comment/drop
On Thu, Sep 01, 2005 at 01:46:00PM -0400, Tom Lane wrote:
> Martijn van Oosterhout writes:
> > 1. Is something supported by glibc usable for us (re portability to
> > non-glibc platforms)?
>
> Nope. Sorry.
Do we have some platforms that don't have any multi-language support? I
mean, we don't ha
"Marc G. Fournier" <[EMAIL PROTECTED]> writes:
> On Mon, 29 Aug 2005, Tom Lane wrote:
>> No, because there's no built-in cast from smallint to bool.
> 'k, I just took a read through the "CREATE CAST" man page, and don't think
> I can use that for this,
Sure you can. Make a SQL or PLPGSQL functi
On Mon, 29 Aug 2005, Tom Lane wrote:
"Marc G. Fournier" <[EMAIL PROTECTED]> writes:
# ALTER TABLE table ALTER COLUMN field1 type boolean;
ERROR: column "field1" cannot be cast to type "pg_catalog.bool"
Should this not work?
No, because there's no built-in cast from smallint to bool. You
Consider the following example:
regression=# create user u1;
CREATE ROLE
regression=# create user u2;
CREATE ROLE
regression=# create user u3;
CREATE ROLE
regression=# grant u1 to u2;
GRANT ROLE
regression=# \c - u1
You are now connected as new user "u1".
regression=> create table t1(f1 int);
CREA
I'm starting to get into PostgreSQL development by implementing:
%Allow ALTER TABLE ... ALTER CONSTRAINT ... RENAME
from the TODO. I've been looking through the code from CommentConstraint
and ATExecDropConstraint and they error out on duplicate constraint
names for a relation. However, A
Darcy Buskermolen <[EMAIL PROTECTED]> writes:
> Yes I've seen the same frustration from users, I think the banner should say
> something along the lines of "psql client version foo connected to server bar
> version sfoo"
That seems overly verbose, particularly in the normal case where the
versio
Alvaro Herrera <[EMAIL PROTECTED]> writes:
> On Thu, Sep 01, 2005 at 04:34:09PM +0100, Simon Riggs wrote:
>> On Wed, 2005-08-31 at 22:25 -0400, Alvaro Herrera wrote:
>>> Now, one thing of note is that you need to "compress" the page in order
>>> to actually be able to use the just-freed space. VAC
Patrick Welche <[EMAIL PROTECTED]> writes:
> I tried the fix mentioned in the earlier message to encourage
> validation. Now dumping this "fixed" database, and loadinging it into the
> new database gives:
> ALTER FUNCTION
> psql:./huge.db:4403: ERROR: function plpgsql_validator(oid) does not exis
Martijn van Oosterhout writes:
> 1. Is something supported by glibc usable for us (re portability to
> non-glibc platforms)?
Nope. Sorry.
regards, tom lane
---(end of broadcast)---
TIP 9: In versions below 8.0, the planner
Matt Miller <[EMAIL PROTECTED]> writes:
> doc/src/sgml/storage.sgml says:
There's a number of things not updated yet in that file :-(
I believe it hasn't heard of pg_twophase either.
regards, tom lane
---(end of broadcast)--
On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
> Andrew Dunstan <[EMAIL PROTECTED]> writes:
>> Tom Lane wrote:
>>> Change the ownership of public in template1 to be a "dbadmin" group.
>>> Grant membership in "dbadmin" to all the DB owners. End of problem.
>
>> Won't that suddenly grant the own
Patrick Welche <[EMAIL PROTECTED]> writes:
> I think we still recommend using *new* pg_dump to dump *old* server when
> upgrading. If one tries that with today's pg_dump (8.1beta1) against
> a 8.1devel server of 6 May, i.e., predating roles, one gets:
[ shrug... ] The current code is expecting t
Simon Riggs <[EMAIL PROTECTED]> writes:
> My first proposal is to add an extra parameter onto the
> index_bulk_delete() call - ntuples. If ntuples == 0 then btbulkdelete()
> will avoid scanning and return immediately. If a scan occurs, then we
> keep track of how many tuples have been marked delete
Josh Berkus writes:
>> THe only fundamental disadvantage that COPY labors under is having to
>> write WAL records. It might be interesting to do something similar to
>> the recent hacks for CREATE TABLE AS, so that a COPY into a table just
>> created in the current transaction would skip writing
On Thu, Sep 01, 2005 at 04:34:09PM +0100, Simon Riggs wrote:
> On Wed, 2005-08-31 at 22:25 -0400, Alvaro Herrera wrote:
> > I was thinking that when the tuple
> > needs to be obsoleted it would need to grow to accomodate the Xmax, but
> > you are not actually proposing to remove that, so it seems
On Thu, Sep 01, 2005 at 03:36:12PM +, Matt Miller wrote:
> But src/include/storage/bufpage.h says:
>
> "/*
> * Page layout version number 0 is for pre-7.3 Postgres releases.
> * Releases 7.3 and 7.4 use 1, denoting a new HeapTupleHeader layout.
> * Release 8.0 changed the HeapTupleHeader l
On Thu, Sep 01, 2005 at 09:20:48AM -0700, Josh Berkus wrote:
> > What I'm saying is that you can write a heap file, on which the tuples
> > would all have xmin=FrozenTransactionId, xmax=Invalid, and the
> > corresponding bits set in the infomask. This ensures that no matter the
> > state of the s
On Thursday 01 September 2005 08:30, Peter Eisentraut wrote:
> A release or two ago we added the version number to the psql welcome
> banner. I noticed that quite a few people interpret that as the server
> version. Somehow, the explicit display of the version numbers leads them to
> make inference
Peter,
> A release or two ago we added the version number to the psql welcome
> banner. I noticed that quite a few people interpret that as the server
> version. Somehow, the explicit display of the version numbers leads them to
> make inferences that they would otherwise not bother about. Has an
On Thu, Sep 01, 2005 at 05:30:34PM +0200, Peter Eisentraut wrote:
> A release or two ago we added the version number to the psql welcome banner.
> I noticed that quite a few people interpret that as the server version.
> Somehow, the explicit display of the version numbers leads them to make
>
On Thu, Sep 01, 2005 at 09:17:38AM +0800, William ZHANG wrote:
> > Dave Page wrote:
> > >
> > >>* Compile with MSVC on Win32 platforms. MySQL support it.
> > >>
> > >So what? It would take a major amount of work, with no useful benefits.
> >
> > ... and you can compile all the client and libra
Andrew Dunstan <[EMAIL PROTECTED]> writes:
> Tom Lane wrote:
>> Change the ownership of public in template1 to be a "dbadmin" group.
>> Grant membership in "dbadmin" to all the DB owners. End of problem.
> Won't that suddenly grant the owner of foo_db dbadmin rights in bar_db?
> That seems to v
Tom,
> THe only fundamental disadvantage that COPY labors under is having to
> write WAL records. It might be interesting to do something similar to
> the recent hacks for CREATE TABLE AS, so that a COPY into a table just
> created in the current transaction would skip writing WAL and instead
> f
I think we still recommend using *new* pg_dump to dump *old* server when
upgrading. If one tries that with today's pg_dump (8.1beta1) against
a 8.1devel server of 6 May, i.e., predating roles, one gets:
pg_dump: SQL command failed
pg_dump: Error message from server: ERROR: relation "pg_catalog.p
Tom Lane wrote:
Change the ownership of public in template1 to be a "dbadmin" group.
Grant membership in "dbadmin" to all the DB owners. End of problem.
Won't that suddenly grant the owner of foo_db dbadmin rights in bar_db?
That seems to violate the principle of least surprise.
ch
On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
> Andrew - Supernews <[EMAIL PROTECTED]> writes:
>> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>>> Change the ownership of public in template1 to be a "dbadmin" group.
>>> Grant membership in "dbadmin" to all the DB owners. End of problem.
doc/src/sgml/storage.sgml says:
"The last 2 bytes of the page header,
pd_pagesize_version, store both the page size
and a version indicator. Beginning with
PostgreSQL 8.0 the version number is 2;
PostgreSQL 7.3 and 7.4 used version number 1;
prior releases used version number 0."
But src/include
Martijn van Oosterhout writes:
> On Thu, Sep 01, 2005 at 08:50:27AM -0400, Tom Lane wrote:
>> varchar could do something like using 24 bits for the length
>> and 8 bits for an encoded indication of the charset.
> With the unfortunate effect that strings are limited to 16Mb instead of
> 1Gb.
No,
You current problem has to do with multiline "gcc --version". You could
backport the fix for that, or you could wrap gcc with a script so it
only reports only line for --version.
I have 7.1.3 running here, I patched configure as attached. Very ugly,
but it works.
> Easy there guys! I'm in the
Alvaro,
> What issues did he raise on this?
On having no Xmin.
> What I'm saying is that you can write a heap file, on which the tuples
> would all have xmin=FrozenTransactionId, xmax=Invalid, and the
> corresponding bits set in the infomask. This ensures that no matter the
> state of the serve
On Thu, 1 Sep 2005, Tom Lane wrote:
> "Sergey E. Koposov" <[EMAIL PROTECTED]> writes:
> > 2660c2660
> > < if (isdigit(hex))
> > ---
> > > if (isdigit((unsigned char)hex))
>
> Sigh. We keep fixing these, and they keep creeping back in. I wish
> there were a way to get some more-mainstr
On Thu, 2005-09-01 at 10:29 +0100, Simon Riggs wrote:
> On Wed, 2005-08-31 at 22:21 -0400, Tom Lane wrote:
> > Simon Riggs <[EMAIL PROTECTED]> writes:
> > > On Wed, 2005-08-31 at 19:24 -0400, Tom Lane wrote:
> > >> If you don't remove any tuples,
> > >> you don't scan the indexes anyway IIRC.
> >
On Thu, Sep 01, 2005 at 11:08:36AM -0400, Tom Lane wrote:
> Alvaro Herrera <[EMAIL PROTECTED]> writes:
> > What I'm saying is that you can write a heap file, on which the tuples
> > would all have xmin=FrozenTransactionId, xmax=Invalid, and the
> > corresponding bits set in the infomask. This ensu
Browsing the glibc stuff for locales I noticed that glibc does actually
allow you to specify the collation order to strcoll and friends. The
feature is however marked with:
Attention: all these functions are *not* standardized in any form.
This is a proof-of-concept implementation.
They do
A release or two ago we added the version number to the psql welcome banner.
I noticed that quite a few people interpret that as the server version.
Somehow, the explicit display of the version numbers leads them to make
inferences that they would otherwise not bother about. Has anyone else
Andrew - Supernews <[EMAIL PROTECTED]> writes:
> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>> Change the ownership of public in template1 to be a "dbadmin" group.
>> Grant membership in "dbadmin" to all the DB owners. End of problem.
> Only if all db owners are equivalent.
If you don't
On Thu, Sep 01, 2005 at 08:50:27AM -0400, Tom Lane wrote:
> Martijn van Oosterhout writes:
> > Simply pass the (Node*) from the parser and let the function sort it
> > out itself. Except now they have to be written in C. Is this
> > unreasonable,
>
> Nope. You're not going to be writing any inte
William wrote:
> You are right. What I want is VC++ projects(*.dsp, *.dsw). Once the
> project files is created, the maintance work is simply add/remove some
> new/deleted source files (*.c only) from the dsps.
>
> And I think VC++ 6.0 is ok, it is power enough and not so big for
pgsql's
> develop
On Wed, 2005-08-31 at 22:25 -0400, Alvaro Herrera wrote:
> On Thu, Sep 01, 2005 at 10:45:44AM +0900, ITAGAKI Takahiro wrote:
>
> Hi,
>
> > I think it would be a waste to retain xmin and cmin for frozen tuples
> > because their values represent only 'visible for all transactions'.
> > Additionally
Alvaro Herrera <[EMAIL PROTECTED]> writes:
> What I'm saying is that you can write a heap file, on which the tuples
> would all have xmin=FrozenTransactionId, xmax=Invalid, and the
> corresponding bits set in the infomask. This ensures that no matter the
> state of the server, you can plug the fil
On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
> Andrew - Supernews <[EMAIL PROTECTED]> writes:
>> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>>> There is some merit in the thought that the DB owner should be able to
>>> grant and revoke access on the public schema, but that no longer
>
"Sergey E. Koposov" <[EMAIL PROTECTED]> writes:
> 2660c2660
> < if (isdigit(hex))
> ---
> > if (isdigit((unsigned char)hex))
Sigh. We keep fixing these, and they keep creeping back in. I wish
there were a way to get some more-mainstream compiler to warn about
passing chars to the fu
- Original Message -
From: "Andrew Dunstan" <[EMAIL PROTECTED]>
To: "Dave Page"
Cc: "William ZHANG" <[EMAIL PROTECTED]>;
Sent: Wednesday, August 31, 2005 10:24 PM
Subject: Re: [HACKERS] Call for 7.5 feature completion
> Dave Page wrote:
> >
> >>* Compile with MSVC on Win32 platforms.
On Thu, Sep 01, 2005 at 10:15:19AM -0400, Robert Treat wrote:
> On Thursday 01 September 2005 09:15, Tom Lane wrote:
> > Gavin Sherry <[EMAIL PROTECTED]> writes:
> > > And this is the reason why it is old. You're asking for pain if
> > > you're using 7.0.3 and this is just the start.
> >
> > There
Andrew - Supernews <[EMAIL PROTECTED]> writes:
> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>> There is some merit in the thought that the DB owner should be able to
>> grant and revoke access on the public schema, but that no longer
>> requires ownership, only membership in an appropriate
- Original Message -
> From: "Dave Page"
> To: "Andrew Dunstan" <[EMAIL PROTECTED]>; "William ZHANG" <[EMAIL PROTECTED]>
> Cc:
> Sent: Thursday, September 01, 2005 3:21 PM
> Subject: RE: [HACKERS] Call for 7.5 feature completion
> > And even those are a royal pain to maintain, never mi
Hello,
First, I'll show the warnings seen when compiling postgres on
SunOS 5.6 with gcc 3.2.1
copy.c: In function `GetDecimalFromHex':
copy.c:2660: warning: subscript has type `char'
copy.c: In function `CopyReadAttributesText':
copy.c:2805: warning: subscript has type `char'
copy.c:2813: warni
On Thursday 01 September 2005 09:15, Tom Lane wrote:
> Gavin Sherry <[EMAIL PROTECTED]> writes:
> > And this is the reason why it is old. You're asking for pain if you're
> > using 7.0.3 and this is just the start.
>
> There are several good reasons why the community stopped supporting
> versions b
On Wed, Aug 31, 2005 at 09:14:42PM -0700, Josh Berkus wrote:
> > One thing that comes to mind is that this makes somewhat easier to build
> > a tool to write pre-built tables, for bulk-loading purposes. You just
> > construct the binary file with the HEAP_FROZEN bit set, and then attach
> > the f
On Thu, Sep 01, 2005 at 04:21:58AM -, Andrew - Supernews wrote:
> On 2005-09-01, Alvaro Herrera <[EMAIL PROTECTED]> wrote:
> > On Thu, Sep 01, 2005 at 01:57:02AM +0100, Simon Riggs wrote:
> >
> >> > If you're using autovacuum then the problem is already taken care of.
> >>
> >> autovacuum will
Josh Berkus writes:
> So, will per-table XID tracking allow us to avoid *ever* vacuuming some
> tables?
If your definition of "ever" is less than a billion transactions, sure.
(As Simon points out, with time-partitioned data sets that could often
be arranged, so it's not a completely silly resp
On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
> Andrew - Supernews <[EMAIL PROTECTED]> writes:
>> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>>> There's considerable feeling that that TODO item is bogus anyway.
>
>> The issue that I've seen is that currently, allowing non-superusers to
> > The bgwriter could "update" all columns of dead heap tuples in heap
> > pages to NULL and thus also gain free space without the need to
touch
> > the indexes.
> > The slot would stay used but it would need less space.
>
> Not unless it's running a transaction (consider TOAST updates).
Ok,
ITAGAKI Takahiro <[EMAIL PROTECTED]> writes:
> I agree. I think an good position of freezer is on bgwriter.
> My idea is:
> 1. Just before bgwriter writes an dirty page in LRU order,
> 2. Freeze tuples in the page and repair fragmentation.
> 3. (Replace the fsm page that has least freespace.)
Gavin Sherry <[EMAIL PROTECTED]> writes:
> And this is the reason why it is old. You're asking for pain if you're
> using 7.0.3 and this is just the start.
There are several good reasons why the community stopped supporting
versions before 7.2, which we did quite some time ago. Don't expect
a lot
Andrew - Supernews <[EMAIL PROTECTED]> writes:
> On 2005-09-01, Tom Lane <[EMAIL PROTECTED]> wrote:
>> There's considerable feeling that that TODO item is bogus anyway.
> The issue that I've seen is that currently, allowing non-superusers to
> create databases in a useful manner requires all sorts
Martijn van Oosterhout writes:
>TYPMODFUNC =3D function( internal [, sometype ] ) RETURNS int32 or intar=
> ray
> Simply pass the (Node*) from the parser and let the function sort it
> out itself. Except now they have to be written in C. Is this
> unreasonable,
Nope. You're not going to be
"Zeugswetter Andreas DAZ SD" <[EMAIL PROTECTED]> writes:
> The bgwriter could "update" all columns of dead heap tuples in heap
> pages
> to NULL and thus also gain free space without the need to touch the
> indexes.
> The slot would stay used but it would need less space.
Not unless it's running a
> > My wild guess is that deleting all index pointers for a removed
index
> > is more-or-less the same cost as creating new ones for
> > inserted/updated page.
>
> Only if you are willing to make the removal process
> recalculate the index keys from looking at the deleted tuple.
The bgwriter
On Thu, Sep 01, 2005 at 11:12:26AM +0300, Hannu Krosing wrote:
> Maybe make the last one "WITH CHARACTER SET xxx" and promote WITH to a
> real keyword.
>
> It seems a good idea to have WITH as a real keyword anyway, as at least
> ANSI/ISO syntax for recursive queries seem to require it too.
Sorry
On Thu, Sep 01, 2005 at 11:18:04AM +0200, Dennis Bjorklund wrote:
> String types have 3 modifiers, the length, the charset and the collation.
> The syntax of these are defined by the standard so at least that syntax
> ought to be allowed (even if there are more work to actually do anything
> wit
On Wed, 2005-08-31 at 22:21 -0400, Tom Lane wrote:
> Simon Riggs <[EMAIL PROTECTED]> writes:
> > On Wed, 2005-08-31 at 19:24 -0400, Tom Lane wrote:
> >> If you don't remove any tuples,
> >> you don't scan the indexes anyway IIRC.
>
> > No. Even if you remove *zero* tuples, an index is still scanne
On Thu, 1 Sep 2005, Martijn van Oosterhout wrote:
> Err, well. My thought was a certain group of type-suffix options would
> be permitted (only zero or one at a time), for example:
>
>WITH TIME ZONE
>WITHOUT TIME ZONE
>CHARACTER SET xxx
String types have 3 modifiers, the length, the
On N, 2005-09-01 at 09:26 +0200, Martijn van Oosterhout wrote:
> On Wed, Aug 31, 2005 at 05:14:13PM -0400, Tom Lane wrote:
> > That strikes me as an unnecessary reduction in flexibility. As long as
> > we make the hardwired type names translate to qualified names (same as
> > they do now) we don't
On Wed, Aug 31, 2005 at 05:14:13PM -0400, Tom Lane wrote:
> That strikes me as an unnecessary reduction in flexibility. As long as
> we make the hardwired type names translate to qualified names (same as
> they do now) we don't have to assume any such thing.
Ack, there's fortunatly only a handful
> -Original Message-
> From: Andrew Dunstan [mailto:[EMAIL PROTECTED]
> Sent: 01 September 2005 03:31
> To: William ZHANG
> Cc: Dave Page; pgsql-hackers@postgresql.org
> Subject: Re: [HACKERS] Call for 7.5 feature completion
>
>
> We currently have nmake files for the client libraries
99 matches
Mail list logo
