I thought we had sufficiently destroyed that reuse a tuple
meme yesterday. You can't do that: there are too many
aspects of the system design that are predicated on the
assumption that dead tuples do not come back to life. You
have to do the full vacuuming bit (index entry removal,
Ühel kenal päeval, N, 2006-03-02 kell 09:53, kirjutas Zeugswetter
Andreas DCP SD:
I thought we had sufficiently destroyed that reuse a tuple
meme yesterday. You can't do that: there are too many
aspects of the system design that are predicated on the
assumption that dead tuples do not
[sorry to everyone if that mail arrives multiple times, but i had
some odd problems with my mail gateway yesterday...]
On Wed, Mar 01, 2006 at 12:41:01PM -0500, Tom Lane wrote:
Peter Eisentraut [EMAIL PROTECTED] writes:
Tom Lane wrote:
How does an optimistic FSM entry avoid the need to run
Peter,
I'd need an invitation to get a visa. Is't possible ?
Oleg
On Wed, 1 Mar 2006, Peter Eisentraut wrote:
PostgreSQL Anniversary Summit
=
Call for Contributions
--
The PostgreSQL Anniversary Summit will take place on July 8 and 9,
On Wed, Mar 01, 2006 at 12:41:01PM -0500, Tom Lane wrote:
Peter Eisentraut [EMAIL PROTECTED] writes:
Tom Lane wrote:
How does an optimistic FSM entry avoid the need to run vacuum?
It ensures that all freed tuples are already in the FSM.
That has nothing to do with it, because the space
This may be a newbie question, but according to the 7.4 docs, an ACCESS
EXCLUSIVE lock is only acquired by the ALTER TABLE, DROP TABLE,
REINDEX, CLUSTER, and VACUUM FULL commands.
However, when viewing pg_locks during the execution of a stored
procedure that does not perform any of the above
Centuries ago, Nostradamus foresaw when [EMAIL PROTECTED] (Tom Lane) would
write:
I thought we had sufficiently destroyed that reuse a tuple meme
yesterday. You can't do that: there are too many aspects of the system
design that are predicated on the assumption that dead tuples do not
come
[EMAIL PROTECTED] wrote:
This may be a newbie question, but according to the 7.4 docs, an ACCESS
EXCLUSIVE lock is only acquired by the ALTER TABLE, DROP TABLE,
REINDEX, CLUSTER, and VACUUM FULL commands.
However, when viewing pg_locks during the execution of a stored
procedure that does
On 2006-03-02, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
No. Here is the offending SP:
CREATE OR REPLACE FUNCTION update_my_cache() RETURNS void AS '
BEGIN
TRUNCATE TABLE my_cache_table;
TRUNCATE is another command that takes an access exclusive lock.
--
Andrew, Supernews
All
I have determined what is causing the failure. It appears that the
stdout stderr redirection to nul produces the Access is Denied.
message. This is happening even if I type dir nul at the command
prompt! I assume that this re-direction in PostgreSQL is done when
starting postgres.exe as to
On Thu, Mar 02, 2006 at 08:33:46AM -0500, Christopher Browne wrote:
What is unclear to me in the discussion is whether or not this is
invalidating the item on the TODO list...
---
Create a bitmap of pages that need vacuuming
snip
I think this is doable, and not invalidated
On 2006-03-02, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
TRUNCATE is another command that takes an access exclusive lock.
The whole SP takes about 10 seconds to run total. The TRUNCATE command
only takes less than a second. However, the access exclusive lock is
held throughout the entire
Mark Dilger [EMAIL PROTECTED] writes:
Tom Lane wrote:
The problem is that you can't determine what answer justify_days would
give without using the assumption 1 month == 30 days, which is an
assumption that justify_hours must not depend on.
Ahhh. So the fact that justify_days already makes
On 2006-03-02, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
First of all, thank you very much. I changed TRUNCATE to DELETE FROM
and my problem as been fixed.
Is there any way to override that behavior? I know you can explicitly
lock tables, can you explicitly unlock tables?
No.
Just to be
Hannu Krosing [EMAIL PROTECTED] writes:
Ãhel kenal päeval, N, 2006-03-02 kell 09:53, kirjutas Zeugswetter
Andreas DCP SD:
Ok, we cannot reuse a dead tuple. Maybe we can reuse the space of a dead
tuple by reducing the tuple to it's header info.
I don't even think you need the header, just
Bernd Helmle [EMAIL PROTECTED] writes:
But couldn't such an opportunistic approach be used for
another lightweight VACUUM mode in such a way, that VACUUM could
look at a special Hot Spot queue, which represents potential
candidates for freeing?
The proposed dirty-page bit map seems a
Cool...
I'm looking forward to 8.2... Ever since 8.x, it seems like we're
progressing rather quickly on performance enhancements, fixes, and new
additions.
On 3/2/06, Tom Lane [EMAIL PROTECTED] wrote:
Robert Treat [EMAIL PROTECTED] writes: I'm trying to plan out a few postgresql related projects
Christopher Browne [EMAIL PROTECTED] writes:
What is unclear to me in the discussion is whether or not this is
invalidating the item on the TODO list...
No, I don't think any of this is an argument against the
dirty-page-bitmap idea. The amount of foreground effort needed to set a
dirty-page
Tom Lane wrote:
Christopher Browne [EMAIL PROTECTED] writes:
What is unclear to me in the discussion is whether or not this is
invalidating the item on the TODO list...
No, I don't think any of this is an argument against the
dirty-page-bitmap idea. The amount of foreground effort needed
On Thu, 2006-03-02 at 00:45, Hannu Krosing wrote:
Ühel kenal päeval, K, 2006-03-01 kell 14:36, kirjutas Scott Marlowe:
But it isn't '-2 months, -1 day'. I think what you are saying is what I
am saying, that we should make the signs consistent.
Pretty much. It just seems wrong to
On Wed, Mar 1, 2006 at 11:02 am, in message
[EMAIL PROTECTED], Simon Riggs
[EMAIL PROTECTED] wrote:
On Wed, 2006- 03- 01 at 10:22 - 0600, Kevin Grittner wrote:
On Tue, Feb 28, 2006 at 7:22 am, in message
[EMAIL PROTECTED], Simon Riggs
[EMAIL PROTECTED] wrote:
We use serializable
Christopher Browne wrote:
What is unclear to me in the discussion is whether or not this is
invalidating the item on the TODO list...
---
Create a bitmap of pages that need vacuuming
Instead of sequentially scanning the entire table, have the background
writer or some
What bothers me about the TODO item is that if we have to sequentially
scan indexes, are we really gaining much by not having to sequentially
scan the heap? If the heap is large enough to gain from a bitmap, the
index is going to be large too. Is disabling per-index cleanout for
expression
All,
This is only the current patch updated to apply cleanly on cvs... it's
not ready for -patches yet as I still haven't spent much time looking
through it and testing it. This is just for anyone to play with
and find issues.
My focus for the next couple days is on getting INS/UPD/DEL RETURNING
Ühel kenal päeval, N, 2006-03-02 kell 10:13, kirjutas Scott Marlowe:
On Thu, 2006-03-02 at 00:45, Hannu Krosing wrote:
Ühel kenal päeval, K, 2006-03-01 kell 14:36, kirjutas Scott Marlowe:
..
We don't say 42 degrees, -12 minutes when measuring arc, do we? Then
again, maybe some folks do.
Csaba Nagy wrote:
What bothers me about the TODO item is that if we have to sequentially
scan indexes, are we really gaining much by not having to sequentially
scan the heap? If the heap is large enough to gain from a bitmap, the
index is going to be large too. Is disabling per-index
Added to TODO:
* Improve port/qsort() to handle sorts with 50% unique and 50% duplicate
value [qsort]
This involves choosing better pivot points for the quicksort.
---
Dann Corbit wrote:
Chris Browne wrote:
oleg@sai.msu.su (Oleg Bartunov) writes:
I'd need an invitation to get a visa. Is't possible ?
Certainty is difficult to promise, but there is a reasonable
population of relevant people here such that invitations can be
arranged.
I suggest that everyone who needs
On Thu, Mar 02, 2006 at 10:06:26AM -0800, Mark Dilger wrote:
I used grep -R to find all locations where interval_justify_time is
mentioned, and for each one added an analogous entry for my new function
interval_justify. But I get lost where OID=1175 is associated with
I think you must keep the header because the tuple might be
part of an update chain (cf vacuuming bugs we repaired just a
few months ago).
t_ctid is potentially interesting data even in a certainly-dead tuple.
yes, I'd still want to keep the full header.
Andreas' idea is possibly doable
Mark Dilger [EMAIL PROTECTED] writes:
I've written the interval_justify() function but the parser does not know
about
it yet.
The pg_proc change is the only source change you need for that, but
afterwards you need to update the postgres.bki file (handled by make and
make install in
My introsort is almost complete and its the fastest variant of
quicksort I can find, I'll submit it to -patches in the next couple
days as-well.On 3/2/06, Bruce Momjian pgman@candle.pha.pa.us wrote:
Added to TODO:* Improve port/qsort() to handle sorts with 50% unique and 50% duplicatevalue
Zeugswetter Andreas DCP SD [EMAIL PROTECTED] writes:
Why do we not truncate the line pointer array ?
Is it, that vacuum (not the full version) does not move
rows to other pages or slots ? Of course vacuum full could do it,
but I see your point.
We can't reassign tuple TIDs safely except in
Just curious, but what is involved in these invitations? For instance,
is there a limit on # of invitations any one person(?) or company can
issue? Are there any legal implications of issuing such an invitation? I
could imagine some pretty hot water if pre 9/11 someone were to invite
bin
Ühel kenal päeval, N, 2006-03-02 kell 15:35, kirjutas Marc G. Fournier:
Just curious, but what is involved in these invitations? For instance,
is there a limit on # of invitations any one person(?) or company can
issue? Are there any legal implications of issuing such an invitation?
Sure.
Tom Lane wrote:
Bruce Momjian pgman@candle.pha.pa.us writes:
This addresses the technical workings of the various backup systems.
The fact that it is or isn't on-line, hot, continuous, fast, or flexible
is secondary.
To me, the continuous activity is the significant feature of that
On Wed, 2006-03-01 at 11:51 +0100, Peter Eisentraut wrote:
The PostgreSQL Anniversary Summit will take place on July 8 and 9, 2006, in
Toronto, Canada. We are planning for a gathering of about 50 hackers,
contributors, and other friends of the PostgreSQL project to celebrate the
project's
Yea, sure I would like to attend.
---
Neil Conway wrote:
On Wed, 2006-03-01 at 11:51 +0100, Peter Eisentraut wrote:
The PostgreSQL Anniversary Summit will take place on July 8 and 9, 2006, in
Toronto, Canada. We are
On Thu, Mar 02, 2006 at 02:49:13PM -0500, Tom Lane wrote:
Michael Fuhr [EMAIL PROTECTED] writes:
Would it make sense for DROP TYPE to have some kind of limited
cascade so you could drop a type and its I/O functions at the same
time, but still get an error if other objects depend on the
I'm creating user-defined server extensions, written in C per the manual 31.9.
C-Language Functions. Everything works well, but only if I fully link the .so such
that there are *no* unresolved external references at all. Not even the stuff in
libstdc++.a can be left out. I've tried setting
Craig A. James wrote:
I'm creating user-defined server extensions, written in C per the
manual 31.9. C-Language Functions. Everything works well, but only
if I fully link the .so such that there are *no* unresolved external
references at all.
What happens if you don't?
--
Peter Eisentraut
Craig A. James [EMAIL PROTECTED] writes:
I'm creating user-defined server extensions, written in C per the
manual 31.9. C-Language Functions. Everything works well, but only
if I fully link the .so such that there are *no* unresolved external
references at all. Not even the stuff in
Here's my very first proof-of-concept patch to PL/pgSQL to use the
RETURNING syntax... INSERT/UPDATE seem to work fine but I think I've
found an error with DELETE RETURNING though, so it doesn't work
properly just yet. Give this a test if you get a chance.
CREATE SEQUENCE test_id_seq START 1
Jonah H. Harris wrote:
All,
This is only the current patch updated to apply cleanly on cvs... it's not
ready for -patches yet as I still haven't spent much time looking through it
and testing it. This is just for anyone to play with and find issues.
Somebody else already did this in the
On Thu, 2006-03-02 at 17:23 -0500, Jonah H. Harris wrote:
If this is the consensus, then I'm fine with posting to -patches
Yeah, -patches is the right place.
I just want to make sure people are aware of it so it can get tested.
I wouldn't expect a whole lot of testing. The usual process is
Tom Lane wrote:
I'm creating user-defined server extensions, written in C per the
manual 31.9. C-Language Functions. Everything works well, but only
if I fully link the .so such that there are *no* unresolved external
references at all. Not even the stuff in libstdc++.a can be left out.
If
I got an installation problem.There was an old copy of postgres
installed butI want to install the modified postgres. I get problems
related to locating "/usr/local/pgsql/data".I tried:[EMAIL PROTECTED]
~]$ /usr/local/pgsql/bin/initdb -D /usr/local/pgsql/dataThe files belonging
to this
FOlks,
One thing I'd like to add: we're considering organizing a code sprint
for the days immediately following the conference.
To add further. There will probably be a code sprint AT the conference
as well. Then Monday and Tuesday for an extended code sprint. We're
still discussing
John [EMAIL PROTECTED] writes:
initializing pg_depend ... sh: line 1: 24405 Segmentation fault=20
/usr/local/pgsql/bin/postgres -F -O -c search_path=3Dpg_catalog -c=20
exit_on_error=3Dtrue template1 /dev/null
Seems like you've got a broken postgres executable there. Where did you
get it from?
George Weaver [EMAIL PROTECTED] writes:
pg_config --version now shows:
$ pg_config --version
PostgreSQL 8.1.0
However, when I try pg_config --pgxs it returns nothing:
$ pg_config --pgxs
$
If you're on Windows this probably means that GetShortPathName() is
failing. I'm not sure what
On Thu, 02 Mar 2006 13:43:35 -0800, Craig A. James wrote:
I'm creating user-defined server extensions, written in C per the manual
[snip]
Is this correct? Do Postgres extension need to be fully statically
linked? Or is there some configuration that will specify LD_LIBRARY_PATH
(or perhaps
Csaba Nagy wrote
From my POV, there must be a way to speed up vacuums on huge tables and
small percentage of to-be-vacuumed tuples... a 200 million rows table
with frequent updates of the _same_ record is causing me some pain right
now. I would like to have that table vacuumed as often as
[ returning to a week-old thread... ]
Stephan Szabo [EMAIL PROTECTED] writes:
On Thu, 23 Feb 2006, Tom Lane wrote:
Any thoughts about details? My feeling is that we should tie RI
semantics to btree opclasses, same as we have done for ORDER BY
and some other SQL constructs, but I don't have a
On Thu, Mar 02, 2006 at 08:05:59PM -0500, Tom Lane wrote:
[ returning to a week-old thread... ]
Stephan Szabo [EMAIL PROTECTED] writes:
On Thu, 23 Feb 2006, Tom Lane wrote:
Any thoughts about details? My feeling is that we should tie RI
semantics to btree opclasses, same as we have done
elein [EMAIL PROTECTED] writes:
... What I'm saying is that the opclass needs to be
an option to PRIMARY KEY and FOREIGN KEY--
PRIMARY KEY and UNIQUE, you mean.
This was brought up before, but I remain less than excited about it.
You can get essentially the same functionality by doing a
The query optimizer currently does not consider reordering a query's
grouping columns. While the order in which ORDER BY columns are
specified affects the semantics of the query, AFAICS GROUP BY's column
order does not. Reordering a query's grouping columns would allow the
optimizer to avoid some
Added to TODO:
* Add missing parameter handling in to_char()
http://archives.postgresql.org/pgsql-hackers/2005-12/msg00948.php
I added a URL in TODO because it is a single message of detail I need to
reference.
The world rejoiced as [EMAIL PROTECTED] (Marc G. Fournier) wrote:
Just curious, but what is involved in these invitations? For
instance, is there a limit on # of invitations any one person(?) or
company can issue? Are there any legal implications of issuing such
an invitation? I could
Added to TODO:
* Allow FSM to return free space toward the beginning of the heap file,
in
hopes that empty pages at the end can be truncated by VACUUM
---
Jim C. Nasby wrote:
On Fri, Dec 09, 2005 at
Is there still interst in this idea for TODO?
---
Jim C. Nasby wrote:
On Thu, Dec 08, 2005 at 10:03:43AM -0500, Bruce Momjian wrote:
Jim C. Nasby wrote:
On Thu, Dec 08, 2005 at 12:59:47AM -0500, Tom Lane wrote:
Christopher Browne [EMAIL PROTECTED] writes:
Letter of Invitation for Countries Whose Citizens Require a
Temporary Resident Visa to Enter Canada
I missed that this was happening up here in Canada. How exclusive is the guest
list for this? Like, are you only expecting 50 top contributors
Where are we on this patch? My testing shows it is still shows we have
a problem:
test= CREATE TABLE x(y INT CHECK(y 0));
CREATE TABLE
test= CREATE TABLE z(a INT) inherits (x);
CREATE TABLE
test= ALTER TABLE z DROP CONSTRAINT x_y_check;
Neil Conway [EMAIL PROTECTED] writes:
A simple hack might help with a subset of this problem, though. For
queries with both ORDER BY and GROUP BY clauses, we can sort the
grouping columns according to their position in the ORDER BY list. So,
given a query like:
SELECT a, b, max(c) FROM t1
On 2/28/06, Suvarna [EMAIL PROTECTED] wrote:
we are using postgresql 7.3.2 version.
As somebody pointed out, that's not a bug... but i think you must
upgrade at least to 7.3.14
even if you really found a bug nobody will fix it for 7.3.2
--
regards,
Jaime Casanova
What they (MySQL) lose in
On Thu, 2 Mar 2006, Tom Lane wrote:
[ returning to a week-old thread... ]
Stephan Szabo [EMAIL PROTECTED] writes:
On Thu, 23 Feb 2006, Tom Lane wrote:
Any thoughts about details? My feeling is that we should tie RI
semantics to btree opclasses, same as we have done for ORDER BY
and
Stephan Szabo [EMAIL PROTECTED] writes:
On Thu, 2 Mar 2006, Tom Lane wrote:
1. If the index opclass contains an exact operator for the case
PKtype = FKtype, use that operator.
Is this rule to read explicitly naming '=' or just the item in that
position in the opclass?
The operator occupying
[EMAIL PROTECTED] (Greg Stark) writes:
Christopher Browne [EMAIL PROTECTED] writes:
Letter of Invitation for Countries Whose Citizens Require a
Temporary Resident Visa to Enter Canada
I missed that this was happening up here in Canada. How exclusive is
the guest list for this? Like, are
btw,
how expensive is to go to the Niagara waterfall from Toronto ?
I'd like to take an opportunity to see it.
Oleg
On Fri, 3 Mar 2006, Chris Browne wrote:
[EMAIL PROTECTED] (Greg Stark) writes:
Christopher Browne [EMAIL PROTECTED] writes:
Letter of Invitation for Countries Whose
Gavin Sherry wrote:
On Wed, 21 Dec 2005, Tom Lane wrote:
Manuel Sugawara masm@fciencias.unam.mx writes:
(Some time ago I proposed an--incomplete--patch and it was rejectd by
Karel arguing that to_char functions should behave *exactly* the same
way that they do in Oracle.)
E - Era name (like, Japanese Imperial) (kind of pointless)
EE - Full era name
Some stuff here:
http://java.sun.com/javase/6/docs/guide/intl/calendar.doc.html
---(end of broadcast)---
TIP 2: Don't 'kill -9' the postmaster
On Fri, 3 Mar 2006 09:07:06 +0300 (MSK)
Oleg Bartunov oleg@sai.msu.su wrote:
how expensive is to go to the Niagara waterfall from Toronto ?
I'd like to take an opportunity to see it.
If you are driving, Niagara Falls is about one hour from Toronto. Cost is a
tank of gas and parking. Looking
Added to TODO:
* Allow to_date() and to_timestamp() accept localized month names
Comment added to the C code to show where it has to happen.
---
Karel Zak wrote:
On Sun, 2005-12-25 at 17:56 -0300, Euler Taveira
72 matches
Mail list logo
