On tis, 2012-05-01 at 20:13 -0400, Tom Lane wrote:
I don't deny that we probably need to reclassify a few error cases,
and fix some elogs that should be ereports, before this approach would
be really workable. My point is that it's *close*, whereas let's
invent some new error severities is
On Tue, 2012-05-01 at 21:22 -0400, David Johnston wrote:
On May 1, 2012, at 20:41, Hannu Krosing ha...@2ndquadrant.com wrote:
Most people don't work in strongly-typed environment, and thus would
work around such restriction if they need a simple JSON value at the
other end of the
On Tue, 2012-05-01 at 19:11 -0400, Tom Lane wrote:
Andrew Dunstan and...@dunslane.net writes:
On Tue, May 1, 2012 at 9:56 AM, Joey Adams
joeyadams3.14...@gmail.comwrote:
No, the RFC says (emphasis mine):
A JSON *text* is a serialized object or array.
If we let the JSON type
I wrote:
Noah Misch wrote:
During ANALYZE, in analyze.c, functions compute_minimal_stats
and compute_scalar_stats, values whose length exceed
WIDTH_THRESHOLD (= 1024) are not used for calculating statistics
other than that they are counted as too wide rows and assumed
to be all different.
On 2012-05-01 22:06, Robert Haas wrote:
It might also be interesting to provide a mechanism to pre-extend a
relation to a certain number of blocks, though if we did that we'd
have to make sure that autovac got the memo not to truncate those
pages away again.
Good point. And just to check
Hi Hackers
How hard would it be to add support for LIKE syntax, similar to table
def in field list declaration for generic record functions
What I'dd like to be able to do is to have a generic json_to_record
function
CREATE OR REPLACE RECORD json_to_record(json) RETURNS RECORD AS $$
...
$$
On 2 May 2012 04:24, Robert Haas robertmh...@gmail.com wrote:
I think Tom's question of whether the parser or lexer is the problem
is something that ought to be investigated. Personally, I suspect
that our tendency to use lists everywhere, for everything, an artifact
of our proud LISP
On Wed, May 2, 2012 at 7:16 AM, Jeroen Vermeulen j...@xs4all.nl wrote:
On 2012-05-01 22:06, Robert Haas wrote:
It might also be interesting to provide a mechanism to pre-extend a
relation to a certain number of blocks, though if we did that we'd
have to make sure that autovac got the memo not
On Thu, Apr 26, 2012 at 8:03 PM, Robert Haas robertmh...@gmail.com wrote:
So, as a first step, I've committed a patch that just throws a hard
conflict. I think we probably want to optimize this further, and I'm
going to work investigate that next. But it seemed productive to get
this much
On Tue, May 1, 2012 at 6:02 PM, Alexander Korotkov aekorot...@gmail.com wrote:
Right. When number of trigrams is big, it is slow to scan posting list of
all of them. The solution is this case is to exclude most frequent trigrams
from index scan. But, it require some kind of statistics of
Peter Eisentraut pete...@gmx.net writes:
On tis, 2012-05-01 at 20:13 -0400, Tom Lane wrote:
I don't deny that we probably need to reclassify a few error cases,
and fix some elogs that should be ereports, before this approach would
be really workable. My point is that it's *close*, whereas
On Wed, May 2, 2012 at 4:50 PM, Robert Haas robertmh...@gmail.com wrote:
On Tue, May 1, 2012 at 6:02 PM, Alexander Korotkov aekorot...@gmail.com
wrote:
Right. When number of trigrams is big, it is slow to scan posting list of
all of them. The solution is this case is to exclude most
On 2 May 2012 04:57, Tom Lane t...@sss.pgh.pa.us wrote:
FWIW, I think only developers not packagers would really be taking such
a hit. I assume we'd continue to ship prebuilt lexer output in
tarballs, so there'd seldom be a reason for a packager to need to run
the tool. Given the extremely
On Wed, May 2, 2012 at 9:35 AM, Alexander Korotkov aekorot...@gmail.com wrote:
I was thinking you could perhaps do it just based on the *number* of
trigrams, not necessarily their frequency.
Imagine we've two queries:
1) SELECT * FROM tbl WHERE col LIKE '%abcd%';
2) SELECT * FROM tbl WHERE
On Wed, May 2, 2012 at 5:48 PM, Robert Haas robertmh...@gmail.com wrote:
On Wed, May 2, 2012 at 9:35 AM, Alexander Korotkov aekorot...@gmail.com
wrote:
Imagine we've two queries:
1) SELECT * FROM tbl WHERE col LIKE '%abcd%';
2) SELECT * FROM tbl WHERE col LIKE '%abcdefghijk%';
The
Tom Lane t...@sss.pgh.pa.us wrote:
My guess is that all the ones defined in the SQL standard are
expected errors, more or less by definition, and thus not
interesting according to Peter G's criteria.
On a scan through the list, I didn't see any exceptions to that,
except for the F0 class.
Kevin Grittner kevin.gritt...@wicourts.gov writes:
That F0 class looks suspicious; are those really defined by standard or
did we encroach on standard naming space with PostgreSQL-specific
values?
I think we screwed up on that :-(. So we ought to renumber those
codes anyway. Perhaps use PF
Tom Lane t...@sss.pgh.pa.us wrote:
Kevin Grittner kevin.gritt...@wicourts.gov writes:
That F0 class looks suspicious; are those really defined by
standard or did we encroach on standard naming space with
PostgreSQL-specific values?
I think we screwed up on that :-(. So we ought to renumber
heap_hot_search_buffer() does this:
valid = HeapTupleSatisfiesVisibility(heapTuple, snapshot, buffer);
If it turns out that the tuple isn't valid (i.e. visible to our scan)
and we haven't yet found any live tuples in the current HOT chain,
then we check whether it's visible to anyone
On Thu, Apr 26, 2012 at 08:56:40AM +0100, Simon Riggs wrote:
A full GTT implementation is not required and the design differed from
that. I don't think hideously complicated is accurate, that's just
you're way of saying and I disagree. Either route is pretty complex
and not much to choose
Attached patch latches up the WAL Writer, reducing wake-ups and thus
saving electricity in a way that is more-or-less analogous to my work
on the BGWriter:
http://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=6d90eaaa89a007e0d365f49d6436f35d2392cfeb
I am hoping this gets into 9.2 . I am
What is the use case for temporary tables on a hot standby server?
Perhaps this is a noobie question, but it seems to me that a hot standby
server's use by* applications* or *users* should be limited to transactions
that don't alter the database in any form.
However, I can see where temporary
On Wed, May 2, 2012 at 11:39 AM, Michael Nolan htf...@gmail.com wrote:
What is the use case for temporary tables on a hot standby server?
Simple...
We required a hot standby server in order to get improved reliability.
But we don't want it to sit there chewing power + money, unused.
We want
Robert Haas robertmh...@gmail.com writes:
+ /* Deleter committed, so tuple is alive if the XID is old enough. */
+ return TransactionIdPrecedes(HeapTupleHeaderGetXmax(tuple), OldestXmin);
s/alive/dead/ in that comment? Otherwise this seems like a good idea.
On Wed, May 2, 2012 at 12:06 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Robert Haas robertmh...@gmail.com writes:
+ /* Deleter committed, so tuple is alive if the XID is old enough. */
+ return TransactionIdPrecedes(HeapTupleHeaderGetXmax(tuple),
OldestXmin);
s/alive/dead/ in that
On Wed, May 2, 2012 at 4:14 PM, Bruce Momjian br...@momjian.us wrote:
I think if implementing global temporary tables only for hot standby
user (#7), it might be of limited usefulness, but the ability to avoid
system table churn (#1) means global temporary tables would have a wide
usefulness,
On Wed, May 2, 2012 at 5:07 PM, Robert Haas robertmh...@gmail.com wrote:
On Wed, May 2, 2012 at 12:06 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Robert Haas robertmh...@gmail.com writes:
+ /* Deleter committed, so tuple is alive if the XID is old enough. */
+ return
Robert Haas robertmh...@gmail.com writes:
On Wed, May 2, 2012 at 12:06 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Otherwise this seems like a good idea.
Do you think I should apply this to 9.2, or wait until 9.3?
We're not at beta yet, and it seems pretty safe/self-contained, so
I have no
On Tue, May 1, 2012 at 5:53 PM, Peter Geoghegan pe...@2ndquadrant.com wrote:
Quite apart from the practical difficulties that we have with Flex
(the fact that the authors are non-responsive and possibly retired,
that annoying compiler warning, and the fact that we are forced to
maintain our
On Sun, Apr 29, 2012 at 10:19:38AM +0100, Simon Riggs wrote:
Maybe we don't need to do this over multiple releases, but we do need
to give warning of possible incompatibilities. It would be good to see
a specific post on hackers called Planned Incompatibilities in 9.2,
or collect such things
Excerpts from Robert Haas's message of mié may 02 08:14:36 -0400 2012:
On Wed, May 2, 2012 at 7:16 AM, Jeroen Vermeulen j...@xs4all.nl wrote:
On 2012-05-01 22:06, Robert Haas wrote:
It might also be interesting to provide a mechanism to pre-extend a
relation to a certain number of blocks,
On 2 May 2012 17:20, Jeff Janes jeff.ja...@gmail.com wrote:
For -S -M simple, the time spent planning is 5 times more than the
time spent parsing. It may be worthwhile to reduce the time spent
parsing, but if the goal is parity with MySQL it probably isn't the
place to start.
Could you
I looked into Maxim Boguk's complaint of bad estimation of antijoin size:
http://archives.postgresql.org/pgsql-general/2012-05/msg00033.php
I can reproduce what I think the problem is in the regression database.
We do okay with this:
regression=# explain analyze select * from tenk1 a where not
On Wed, May 2, 2012 at 12:26 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Excerpts from Robert Haas's message of mié may 02 08:14:36 -0400 2012:
On Wed, May 2, 2012 at 7:16 AM, Jeroen Vermeulen j...@xs4all.nl wrote:
On 2012-05-01 22:06, Robert Haas wrote:
It might also be interesting
On Sun, Apr 29, 2012 at 02:06:48PM -0400, Jay Levitt wrote:
Bruce Momjian wrote:
I am again requesting the addition of options to tools/git_changelog so
I can more easily produce the release notes. I asked for this during
9.1 development and it was rejected. I am currently using my own
Robert Haas robertmh...@gmail.com writes:
On Wed, May 2, 2012 at 12:26 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Hm. I see those two things as different -- to me, bloat is unremoved
dead tuples, whereas slack space would be free space that can be reused
by new tuples. Slack space
Excerpts from Robert Haas's message of mié may 02 12:37:35 -0400 2012:
On Wed, May 2, 2012 at 12:26 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Excerpts from Robert Haas's message of mié may 02 08:14:36 -0400 2012:
On Wed, May 2, 2012 at 7:16 AM, Jeroen Vermeulen j...@xs4all.nl
On Wed, May 2, 2012 at 12:46 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Agreed. Perhaps to solve this issue what we need is a way to migrate
tuples from later pages into earlier ones. (This was one of the points,
never resolved, that we discussed during the VACUUM FULL rework.)
On Wed, May 2, 2012 at 12:19 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Robert Haas robertmh...@gmail.com writes:
On Wed, May 2, 2012 at 12:06 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Otherwise this seems like a good idea.
Do you think I should apply this to 9.2, or wait until 9.3?
We're not at
On ons, 2012-05-02 at 13:40 +0200, Hannu Krosing wrote:
How hard would it be to add support for LIKE syntax, similar to table
def in field list declaration for generic record functions
What I'dd like to be able to do is to have a generic json_to_record
function
CREATE OR REPLACE RECORD
Excerpts from Robert Haas's message of mié may 02 12:55:17 -0400 2012:
On Wed, May 2, 2012 at 12:46 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Agreed. Perhaps to solve this issue what we need is a way to migrate
tuples from later pages into earlier ones. (This was one of the
On ons, 2012-05-02 at 00:31 -0400, Tom Lane wrote:
Peter Eisentraut pete...@gmx.net writes:
Remove dead ports
Checking this patch, I noticed that config.guess and config.sub harbor
most of the remaining references to those platforms, which reminded me:
don't we usually update those files
Peter Eisentraut pete...@gmx.net writes:
On ons, 2012-05-02 at 00:31 -0400, Tom Lane wrote:
Checking this patch, I noticed that config.guess and config.sub harbor
most of the remaining references to those platforms, which reminded me:
don't we usually update those files from autoconf upstream
On Wed, May 2, 2012 at 9:31 AM, Peter Geoghegan pe...@2ndquadrant.com wrote:
On 2 May 2012 17:20, Jeff Janes jeff.ja...@gmail.com wrote:
For -S -M simple, the time spent planning is 5 times more than the
time spent parsing. It may be worthwhile to reduce the time spent
parsing, but if the
On Wed, May 2, 2012 at 3:33 PM, Peter Geoghegan pe...@2ndquadrant.com wrote:
On 2 May 2012 04:57, Tom Lane t...@sss.pgh.pa.us wrote:
FWIW, I think only developers not packagers would really be taking such
a hit. I assume we'd continue to ship prebuilt lexer output in
tarballs, so there'd
Hi Gilles,
Sorry for the delay.
Il 03/04/12 14:21, Gilles Darold ha scritto:
+1, this is also my point of view.
I have looked at the patch that contains both pg_is_in_backup() and
pg_backup_start_time().
From a functional point of view it looks fine to me. I was thinking
of
On lör, 2012-04-28 at 00:32 -0400, Tom Lane wrote:
I'm inclined to think that the best fix is for
PLy_spi_execute_fetch_result to copy the tupledesc into
TopMemoryContext, not the current context. This is a tad scary from a
memory leakage standpoint, but I suppose that if python fails to
On Wed, May 2, 2012 at 1:06 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Excerpts from Robert Haas's message of mié may 02 12:55:17 -0400 2012:
On Wed, May 2, 2012 at 12:46 PM, Alvaro Herrera
alvhe...@commandprompt.com wrote:
Agreed. Perhaps to solve this issue what we need is a way
On Wed, May 02, 2012 at 10:37:58AM -0700, Jeff Janes wrote:
I could try to clean up and post the patch that implements this if you want.
The second method was just to do --enable-profiling on a stock build
and look at the call graph section of gprof output. It attributed 50%
to
On Wed, May 02, 2012 at 01:15:48PM -0400, Tom Lane wrote:
Peter Eisentraut pete...@gmx.net writes:
On ons, 2012-05-02 at 00:31 -0400, Tom Lane wrote:
Checking this patch, I noticed that config.guess and config.sub harbor
most of the remaining references to those platforms, which reminded
This doesn't work anymore with Python 3:
rv = plpy.execute(...)
do_something(rv[0:1])
Apparently, they changed the C API for doing slicing, or rather made one
of the two APIs for it silently do nothing. Details are difficult to
find, but this email message seems to contain something:
On Wed, May 2, 2012 at 12:06 PM, Peter Eisentraut pete...@gmx.net wrote:
On ons, 2012-05-02 at 13:40 +0200, Hannu Krosing wrote:
How hard would it be to add support for LIKE syntax, similar to table
def in field list declaration for generic record functions
What I'dd like to be able to do is
Hello List,
I'd like to share with you some experiences we've had while
investigating what we'd have to do to make very-very tiny databases.
First, the formulae at
http://www.postgresql.org/docs/9.1/static/kernel-resources.html#SHARED-MEMORY-PARAMETERS
(17-2) seem misleading, particularly with
Robert Haas robertmh...@gmail.com writes:
Brainstorming wildly, how about something like this:
1. Insert a new copy of the tuple onto some other heap page. The new
tuple's xmin will be that of the process doing the tuple move, and
we'll also set a flag indicating that a move is in progress.
On Wed, May 2, 2012 at 4:39 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Robert Haas robertmh...@gmail.com writes:
Brainstorming wildly, how about something like this:
1. Insert a new copy of the tuple onto some other heap page. The new
tuple's xmin will be that of the process doing the tuple
Kevin Grittner kevin.gritt...@wicourts.gov writes:
Tom Lane t...@sss.pgh.pa.us wrote:
Kevin Grittner kevin.gritt...@wicourts.gov writes:
That F0 class looks suspicious; are those really defined by
standard or did we encroach on standard naming space with
PostgreSQL-specific values?
I think
On Fri, Mar 23, 2012 at 10:03 AM, Fujii Masao masao.fu...@gmail.com wrote:
On second thought, I found other issues about WAL archiving after
failover. So let me clarify the issues again.
Just after failover, there can be three kinds of WAL files in new
master's pg_xlog directory:
(1) WAL
Peter Geoghegan pe...@2ndquadrant.com writes:
Attached patch latches up the WAL Writer, reducing wake-ups and thus
saving electricity in a way that is more-or-less analogous to my work
on the BGWriter:
On Wed, May 2, 2012 at 2:37 PM, Vivek Singh Raghuwanshi
vivekraghuwan...@gmail.com wrote:
Please send me the link or white papers from where i can get information
like.
3. Feature comparison of PostgreSQL and Postgres-XC
4. and can we use Postgres-XC in production with mission critical env
According to
http://archives.postgresql.org/pgsql-general/2012-04/msg00374.php
advisory locks now cause problems for prepared transactions, which
ought to ignore them. It appears to me that this got broken by
commit 62c7bd31c8878dd45c9b9b2429ab7a12103f3590, which marked the
userlock lock method
On Wed, May 2, 2012 at 7:21 PM, Tom Lane t...@sss.pgh.pa.us wrote:
It is getting a bit late to be considering such changes for 9.2, but
I'm willing to review and commit this if there's not anybody who feels
strongly that it's too late. Personally I think it's in the nature of
cleanup and so
Michael,
What is the use case for temporary tables on a hot standby server?
Perhaps this is a noobie question, but it seems to me that a hot standby
server's use by* applications* or *users* should be limited to transactions
that don't alter the database in any form.
A very common use for
On 5/2/12 10:20 AM, Jameison Martin wrote:
Attached are the following as per various requests:
* test_results.txt: the performance benchmarking results,
* TestTrailingNull.java: the performance benchmarking code, with a few
additional scenarios as per various requests
On Mon, Apr 30, 2012 at 02:35:20PM -0400, Tom Lane wrote:
Noah Misch n...@leadboat.com writes:
When GIN changes a metapage, we WAL-log its ex-header content and never use
a
backup block. This reduces WAL volume since the vast majority of the
metapage
is unused. However,
Robert Haas robertmh...@gmail.com writes:
... It seems unlikely to cause any real
problem if WAL writer takes a couple seconds to get with the program
after a long period of inactivity; note that an async commit will kick
it anyway, and a sync commit will probably half to flush WAL whether
Thanks this is very helpful
Regards
ViVek
On Thu, May 3, 2012 at 4:57 AM, Michael Paquier
michael.paqu...@gmail.comwrote:
On Wed, May 2, 2012 at 2:37 PM, Vivek Singh Raghuwanshi
vivekraghuwan...@gmail.com wrote:
Please send me the link or white papers from where i can get information
On Wed, May 2, 2012 at 6:06 PM, Noah Misch n...@leadboat.com wrote:
Can we indeed assume that all support-worthy filesystems align the start of
every file to a physical sector? I know little about modern filesystem
design, but these references leave me wary of that assumption:
On Wed, May 2, 2012 at 11:42 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Robert Haas robertmh...@gmail.com writes:
... It seems unlikely to cause any real
problem if WAL writer takes a couple seconds to get with the program
after a long period of inactivity; note that an async commit will kick
it
Daniel Farina dan...@heroku.com writes:
On Wed, May 2, 2012 at 6:06 PM, Noah Misch n...@leadboat.com wrote:
Can we indeed assume that all support-worthy filesystems align the start of
every file to a physical sector? I know little about modern filesystem
design, but these references leave me
On 4/29/12 9:27 AM, Kevin Grittner wrote:
Maybe I can help with that by describing what the Wisconsin court
system does for circuit court data.
Thanks for the write-up, it was insightful.
One thing I wanted to mention is that non-binary replication has an added
advantage over binary from a
On 4/29/12 6:03 AM, Simon Riggs wrote:
The DML-WITH-LIMIT-1 is required to do single logical updates on tables
with non-unique rows.
And as for any logical updates we will have huge performance problem
when doing UPDATE or DELETE on large table with no indexes, but
fortunately this
71 matches
Mail list logo