(2012/03/12 19:21), Etsuro Fujita wrote:
According to the following documentation on IterateForeignScan() in
50.2. Foreign Data Wrapper Callback Routines, I have created a patch to
support the error handling in file_fdw. Please find attached a patch.
Note that PostgreSQL's executor
On Tue, Mar 13, 2012 at 2:50 AM, Tom Lane t...@sss.pgh.pa.us wrote:
Info appreciated.
Email seen, will reply when I can later today.
--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training Services
--
Sent via pgsql-hackers mailing list
If we have a query of the form:
Select *, (Select * FROM T2 WHERE p2 = T1.p1) FROM T1 ORDER
BY 1 WHERE p3 = 75
In SQL, if a subselect is present in the column-list, is the
result set of this subselect considered to be a part of a
single tuple returned from the outer query, or does the result
set
On Mon, Mar 12, 2012 at 8:10 PM, Bruce Momjian br...@momjian.us wrote:
To answer your specific question, I think clearing the last analyzed
fields should cause autovacuum to run on analyze those tables. What I
don't know is whether not clearing the last vacuum datetime will cause
the table
On Mon, Mar 12, 2012 at 9:12 PM, Tom Lane t...@sss.pgh.pa.us wrote:
Bruce Momjian br...@momjian.us writes:
Copying the statistics from the old server is on the pg_upgrade TODO
list. I have avoided it because it will add an additional requirement
that will make pg_upgrade more fragile in case
On Tuesday, March 13, 2012 04:49:40 AM Jeff Davis wrote:
On Sun, 2012-02-05 at 17:56 -0500, Noah Misch wrote:
I meant primarily to illustrate the need to be comprehensive, not comment
on which executable should fsync a particular file. Bootstrap-mode
backends do not sync anything during an
On Tuesday, March 13, 2012 03:26:34 AM Robert Haas wrote:
Meanwhile, here are some TPS graphs at 16MB and 32MB on the IBM POWER7
machine. 32 clients, 1800 seconds, scale factor 300, synchronous
commit off.
That graph makes me cringe because its pretty representative of what I have
seen in
(2012/03/13 15:53), Shigeru HANADA wrote:
(2012/03/12 19:21), Etsuro Fujita wrote:
According to the following documentation on IterateForeignScan() in
50.2. Foreign Data Wrapper Callback Routines, I have created a patch to
support the error handling in file_fdw. Please find attached a patch.
Hi,
I did a short review of what I found after merging master
(b4af1c25bbc636379efc5d2ffb9d420765705b8a) to what I currently fetched from
your repo (d63df64580114de4d83cfe8eb45eb630724b8b6f).
- I still find it strange not to fire on cascading actions
- I dislike the missing locking leading to
On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes jeff.ja...@gmail.com wrote:
Rerunning all 4 benchmarks (both 16MB and 32MB wal_buffers on both
machines) with fsync=off (as well as synchronous_commit=off still)
might help clarify things.
If it increases the TPS of Nate@16MB, but doesn't change the
Joachim Wieland j...@mcknight.de writes:
On Sat, Mar 10, 2012 at 9:51 AM, Robert Haas robertmh...@gmail.com wrote:
-const char *owner, bool withOids,
+const char *owner,
+unsigned long int relpages, bool withOids,
The
On Mon, Mar 12, 2012 at 11:10 PM, Jaime Casanova ja...@2ndquadrant.com wrote:
On Mon, Mar 12, 2012 at 9:41 PM, Noah Misch n...@leadboat.com wrote:
I created a CF entry for this and marked it Ready for Committer.
i wasn't sure if create an entry this late was a good idea or not...
but now i
Tom Lane t...@sss.pgh.pa.us wrote:
(I'm also unconvinced that sorting by relation size is a good idea
anyway. Anything that makes the dump order less predictable gets
push-back, IME.)
Given that people often use diff on files from pg_dump,
unpredictable ordering can be a bad thing. On the
On Tuesday, March 13, 2012 02:48:11 PM Tom Lane wrote:
(I'm also unconvinced that sorting by relation size is a good idea
anyway. Anything that makes the dump order less predictable gets
push-back, IME.)
Why? Especially in the directory format - which is a prerequisite for parallel
dump if I
The introduction of custom_variable_classes and #variable_conflict in
9.0 partly solves the problem with mixing IN/OUT variables with column
names.
In 8.4 and before, it defaulted to the IN/OUT variable, if it shared
the name of a column.
In 9.0 the behaviour was changed to raise an error if a
On Tue, Mar 13, 2012 at 12:12:27AM -0400, Tom Lane wrote:
Bruce Momjian br...@momjian.us writes:
Copying the statistics from the old server is on the pg_upgrade TODO
list. I have avoided it because it will add an additional requirement
that will make pg_upgrade more fragile in case of
2012/3/12 Robert Haas robertmh...@gmail.com:
On Mon, Mar 12, 2012 at 12:30 PM, Kohei KaiGai kai...@kaigai.gr.jp wrote:
Suppose that the connection starts out in context connection_pooler_t.
Based on the identity of the user, we transition to foo_t, bar_t, or
baz_t. If it's possible, by any
Hello
you can use function name as qualifier
create or replace function fx(paramname type, ...)
returns ...
begin
SELECT INTO fx.paramname, ...
Regards
Pavel Stehule
2012/3/13 Joel Jacobson j...@trustly.com:
The introduction of custom_variable_classes and #variable_conflict in
9.0 partly
On Tue, Mar 13, 2012 at 12:33:09AM -0700, Daniel Farina wrote:
On Mon, Mar 12, 2012 at 8:10 PM, Bruce Momjian br...@momjian.us wrote:
To answer your specific question, I think clearing the last analyzed
fields should cause autovacuum to run on analyze those tables. What I
don't know is
Joel Jacobson j...@trustly.com writes:
In situations when it's natural and makes sense to mix IN/OUT
variables and columns names, it would be nice to being able to
explicitly specifying you are referring to the IN or OUT variable with
a specific name.
Can't you qualify them with the function
Bruce Momjian br...@momjian.us wrote:
I just received a sobering blog comment stating that pg_upgrade
took 5 minutes on a 0.5TB database, but analyze took over an hour:
Yeah, we have had similar experiences. Even if this can't be done
for every release or for every data type, bringing over
Excerpts from Bruce Momjian's message of mar mar 13 11:14:43 -0300 2012:
On Tue, Mar 13, 2012 at 12:33:09AM -0700, Daniel Farina wrote:
On Mon, Mar 12, 2012 at 8:10 PM, Bruce Momjian br...@momjian.us wrote:
To answer your specific question, I think clearing the last analyzed
fields
On Tue, Mar 13, 2012 at 11:34:16AM -0300, Alvaro Herrera wrote:
Excerpts from Bruce Momjian's message of mar mar 13 11:14:43 -0300 2012:
On Tue, Mar 13, 2012 at 12:33:09AM -0700, Daniel Farina wrote:
On Mon, Mar 12, 2012 at 8:10 PM, Bruce Momjian br...@momjian.us wrote:
To answer your
Excerpts from Bruce Momjian's message of mar mar 13 11:49:26 -0300 2012:
On Tue, Mar 13, 2012 at 11:34:16AM -0300, Alvaro Herrera wrote:
Excerpts from Bruce Momjian's message of mar mar 13 11:14:43 -0300 2012:
On Tue, Mar 13, 2012 at 12:33:09AM -0700, Daniel Farina wrote:
On Mon,
amit sehas cu...@yahoo.com wrote:
If we have a query of the form:
Select *, (Select * FROM T2 WHERE p2 = T1.p1) FROM T1 ORDER
BY 1 WHERE p3 = 75
In SQL, if a subselect is present in the column-list, is the
result set of this subselect considered to be a part of a
single tuple returned
Excerpts from Andres Freund's message of mar mar 13 08:22:26 -0300 2012:
- I think list_command_triggers should do a heap_lock_tuple(LockTupleShared)
on the command trigger tuple. But then again just about nothing else does :(
If you want to do something like that, I think it's probably more
On Tue, Mar 13, 2012 at 09:28:58AM -0500, Kevin Grittner wrote:
Bruce Momjian br...@momjian.us wrote:
I just received a sobering blog comment stating that pg_upgrade
took 5 minutes on a 0.5TB database, but analyze took over an hour:
Yeah, we have had similar experiences. Even if this
On Tue, Mar 13, 2012 at 12:08:41PM -0300, Alvaro Herrera wrote:
You're wrong. Autovacuum does not consider time, only dead/live tuple
counts. The formulas it uses are in the autovacuum docs; some details
(such as the fact that it skips tables that do not have stat entries)
might be
Folks,
This is for 9.3, of course.
I noticed that CREATE FOREIGN TABLE (LIKE some_table) doesn't work. I
believe it should, as it would:
- Remove a POLA violation
- Make data loading into an extant table even easier, especially if
there need to be filtering or other cleanup steps
Come to
On Tue, Mar 06, 2012 at 04:39:32PM -0300, Alvaro Herrera wrote:
Here's a first attempt at a README illustrating this. I intend this to
be placed in src/backend/access/heap/README.tuplock; the first three
paragraphs are stolen from the comment in heap_lock_tuple, so I'd remove
those from
On Mon, Mar 12, 2012 at 9:24 PM, Noah Misch n...@leadboat.com wrote:
When we lock an update-in-progress row, we walk the t_ctid chain and lock all
descendant tuples. They may all have uncommitted xmins. This is essential to
ensure that the final outcome of the updating transaction does not
On Mon, Feb 27, 2012 at 05:26:04PM +0100, Magnus Hagander wrote:
On Mon, Feb 27, 2012 at 16:20, Tom Lane t...@sss.pgh.pa.us wrote:
Magnus Hagander mag...@hagander.net writes:
On Mon, Feb 27, 2012 at 04:37, Robert Haas robertmh...@gmail.com wrote:
Why not change the default? Does anyone
=?UTF-8?B?SmFuIFVyYmHFhHNraQ==?= wulc...@wulczer.org writes:
I came up with a stack of context structures that gets pushed when a
PL/Python starts being executed and popped when it returns. At first
they contained just a scratch memory context used by PLyDict_FromTuple.
Then under the premise
Excerpts from Bruce Momjian's message of mar mar 13 14:00:52 -0300 2012:
On Tue, Mar 06, 2012 at 04:39:32PM -0300, Alvaro Herrera wrote:
When there is a single locker in a tuple, we can just store the locking info
in the tuple itself. We do this by storing the locker's Xid in XMAX, and
On Mon, Mar 12, 2012 at 3:28 PM, Simon Riggs si...@2ndquadrant.com wrote:
I agree with you that some worst case performance tests should be
done. Could you please say what you think the worst cases would be, so
those can be tested? That would avoid wasting time or getting anything
backwards.
On Tue, Mar 13, 2012 at 1:38 AM, Daniel Farina dan...@heroku.com wrote:
You probably are going to ask: why not just run ANALYZE and be done
with it?
Uhm yes. If analyze takes a long time then something is broken. It's
only reading a sample which should be pretty much a fixed number of
pages per
On Mon, Mar 12, 2012 at 11:35 PM, Joachim Wieland j...@mcknight.de wrote:
How do you mean it's unused? pg_dump_sort.c uses relpages to dump the
largest tables first. What you don't want to see in a parallel dump is
a worker starting to dump a large table while everybody else is
already idle...
On 03/13/2012 01:53 PM, Robert Haas wrote:
I tried this actually (patch attached) but then I wanted to test it
and couldn't find anything that used pgpipe() on Windows.
pg_basebackup/pg_basebackup.c is using it but it's in an #ifndef WIN32
and the same is true for postmaster/syslogger.c. Am
Greg Stark st...@mit.edu wrote:
Daniel Farina dan...@heroku.com wrote:
You probably are going to ask: why not just run ANALYZE and be
done with it?
Uhm yes. If analyze takes a long time then something is broken.
It's only reading a sample which should be pretty much a fixed
number of pages
On sön, 2012-02-19 at 20:23 +0100, Pavel Stehule wrote:
other very simple patch - enhance autocomplete to support CREATE OR
REPLACE FUNCTION statement
I see two problems with this:
- We also have other CREATE OR REPLACE variants that this won't
complete.
- It won't complete something like
On Tue, Mar 13, 2012 at 05:46:06PM +, Greg Stark wrote:
On Tue, Mar 13, 2012 at 1:38 AM, Daniel Farina dan...@heroku.com wrote:
You probably are going to ask: why not just run ANALYZE and be done
with it?
Uhm yes. If analyze takes a long time then something is broken. It's
only
On Tue, Mar 13, 2012 at 01:18:58PM -0500, Kevin Grittner wrote:
Greg Stark st...@mit.edu wrote:
Daniel Farina dan...@heroku.com wrote:
You probably are going to ask: why not just run ANALYZE and be
done with it?
Uhm yes. If analyze takes a long time then something is broken.
It's
On Wed, Feb 29, 2012 at 01:27:43PM -0500, Robert Haas wrote:
On Wed, Feb 29, 2012 at 12:20 PM, Thom Brown t...@linux.com wrote:
On 29 February 2012 17:16, Tom Lane t...@sss.pgh.pa.us wrote:
Thom Brown t...@linux.com writes:
So could we introduce either a command to show which objects are
On 2 March 2012 20:10, Tom Lane t...@sss.pgh.pa.us wrote:
I do intend to take this one up in due course
I probably should have exposed the query_id directly in the
pg_stat_statements view, perhaps as query_hash. The idea of that
would be to advertise the potential non-uniqueness of the value - a
Bruce Momjian br...@momjian.us wrote:
On Tue, Mar 13, 2012 at 01:18:58PM -0500, Kevin Grittner wrote:
cir=# analyze CaseHist;
ANALYZE
Time: 143450.467 ms
cir=# select relpages, reltuples from pg_class where relname =
'CaseHist';
relpages | reltuples
--+-
On Tue, Mar 13, 2012 at 9:59 AM, Kevin Grittner
kevin.gritt...@wicourts.gov wrote:
Tom Lane t...@sss.pgh.pa.us wrote:
(I'm also unconvinced that sorting by relation size is a good idea
anyway. Anything that makes the dump order less predictable gets
push-back, IME.)
Given that people often
On Tue, Mar 13, 2012 at 2:56 PM, Bruce Momjian br...@momjian.us wrote:
Is this a TODO?
If you're going to create one, link to Álvaro's message. But I'm not
sure we should without a better sense of what we actually want to do,
from the options he laid out.
--
Robert Haas
EnterpriseDB:
Kevin Grittner kevin.gritt...@wicourts.gov wrote:
Bruce Momjian br...@momjian.us wrote:
That is 2.5 minutes. How large is that database?
I dug around a little and found that we had turned on vacuum cost
limits on the central databases, because otherwise the web team
complained about
=?UTF-8?B?SmFuIFVyYmHFhHNraQ==?= wulc...@wulczer.org writes:
Here are the updated patches which use PLy_elog instead of plain elog.
The difference is that they will get marked for translation and that the
original Python exception will show up in the errdetail field.
Applied, thanks.
On Tue, Mar 13, 2012 at 02:07:14PM -0500, Kevin Grittner wrote:
Bruce Momjian br...@momjian.us wrote:
On Tue, Mar 13, 2012 at 01:18:58PM -0500, Kevin Grittner wrote:
cir=# analyze CaseHist;
ANALYZE
Time: 143450.467 ms
cir=# select relpages, reltuples from pg_class where relname =
On lör, 2012-03-10 at 18:47 -0500, Tom Lane wrote:
void EncodeDateTime(struct pg_tm * tm, fsec_t fsec, const int *tzp, const
char *tzn, int style, char *str)
It appears to me that null-ness of tzp and tzn are used as a 3-way flag
to identify the style of timezone output wanted (none,
Greg Stark st...@mit.edu writes:
On Tue, Mar 13, 2012 at 1:38 AM, Daniel Farina dan...@heroku.com wrote:
You probably are going to ask: why not just run ANALYZE and be done
with it?
Uhm yes. If analyze takes a long time then something is broken. It's
only reading a sample which should be
On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes jeff.ja...@gmail.com wrote:
Rerunning all 4 benchmarks (both 16MB and 32MB wal_buffers on both
machines) with fsync=off (as well as synchronous_commit=off still)
might help clarify things.
I reran the 32-client benchmark on the IBM machine with
Bruce Momjian br...@momjian.us wrote:
cir=# analyze CaseHist;
ANALYZE
Time: 143450.467 ms
OK, so a single 44GB tables took 2.5 minutes to analyze; that is
not good. It would require 11 such tables to reach 500GB (0.5
TB), and would take 27 minutes. The report I had was twice as
long,
Peter Geoghegan pe...@2ndquadrant.com writes:
I probably should have exposed the query_id directly in the
pg_stat_statements view, perhaps as query_hash.
FWIW, I think that's a pretty bad idea; the hash seems to me to be
strictly an internal matter. Given the sponginess of its definition
I
Hi,
Andres Freund and...@anarazel.de writes:
I did a short review of what I found after merging master
Thanks!
- I still find it strange not to fire on cascading actions
We don't build statement for cascading so we don't fire command
triggers. The user view is that there was no drop command
On tis, 2012-03-13 at 15:44 -0400, Tom Lane wrote:
I wonder whether it'd be worth recommending that people do an initial
ANALYZE with a low stats target, just to get some stats in place,
and then go back to analyze at whatever their normal setting is.
Perhaps going even further, ANALYZE could
Bruce Momjian br...@momjian.us wrote:
OK, so a single 44GB tables took 2.5 minutes to analyze; that is
not good. It would require 11 such tables to reach 500GB (0.5
TB), and would take 27 minutes. The report I had was twice as
long, but still in the ballpark of too long. :-(
But it's
On Tue, Mar 13, 2012 at 10:10:02PM +0200, Peter Eisentraut wrote:
On tis, 2012-03-13 at 15:44 -0400, Tom Lane wrote:
I wonder whether it'd be worth recommending that people do an initial
ANALYZE with a low stats target, just to get some stats in place,
and then go back to analyze at
Bruce Momjian br...@momjian.us wrote:
OK, good idea. Kevin, can you test this:
PGOPTIONS='-c default_statistics_target=10' vacuumdb --all
--analyze-only
Is it faster? Thanks.
Well, I just did something similar in psql -- I disabled the delays
by:
set vacuum_cost_delay = 0;
On Tuesday, March 13, 2012 09:07:32 PM Dimitri Fontaine wrote:
Hi,
Andres Freund and...@anarazel.de writes:
I did a short review of what I found after merging master
Thanks!
- I still find it strange not to fire on cascading actions
We don't build statement for cascading so we
Peter Eisentraut pete...@gmx.net writes:
On lör, 2012-03-10 at 18:47 -0500, Tom Lane wrote:
It appears to me that null-ness of tzp and tzn are used as a 3-way flag
to identify the style of timezone output wanted (none, numeric, or alpha).
It's not quite a three-way flag, because it also
On Tue, Mar 13, 2012 at 7:30 PM, Bruce Momjian br...@momjian.us wrote:
OK, so a single 44GB tables took 2.5 minutes to analyze; that is not
good. It would require 11 such tables to reach 500GB (0.5 TB), and
would take 27 minutes. The report I had was twice as long, but still in
the ballpark
On Tue, Mar 13, 2012 at 09:28:33PM +, Greg Stark wrote:
hmph. One thing that could speed up analyze on raid arrays would be
doing prefetching so more than one spindle can be busy. Sacrificing
statistical accuracy by reading a less random sample on contiguous
blocks of rows would also be
On Tue, Mar 13, 2012 at 03:29:22PM -0500, Kevin Grittner wrote:
I went even lower than you suggested:
set default_statistics_target = 4;
And it was much faster, but still more time than the pg_upgrade run
itself:
cir=# analyze;
ANALYZE
Time: 474319.826 ms
A little under 8
On Tue, Mar 13, 2012 at 3:48 PM, Robert Haas robertmh...@gmail.com wrote:
On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes jeff.ja...@gmail.com wrote:
Rerunning all 4 benchmarks (both 16MB and 32MB wal_buffers on both
machines) with fsync=off (as well as synchronous_commit=off still)
might help
On Tue, Mar 13, 2012 at 4:55 AM, Andres Freund and...@anarazel.de wrote:
On Tuesday, March 13, 2012 03:26:34 AM Robert Haas wrote:
Meanwhile, here are some TPS graphs at 16MB and 32MB on the IBM POWER7
machine. 32 clients, 1800 seconds, scale factor 300, synchronous
commit off.
That graph
On Tue, Mar 13, 2012 at 5:42 PM, Bruce Momjian br...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run vacuumdb twice, once
with default_statistics_target = 4, and another with the default?
I'm not sure why
Bruce Momjian br...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run vacuumdb
twice, once with default_statistics_target = 4, and another with
the default?
Here are the results at various settings.
1 :
Hi,
Has anyone ever complained about the inconsistency (at least it seems
that to me) of using INHERITS in CREATE TABLE and INHERIT in ALTER
TABLE?
Anyone, besides me, think we should fix that? Maybe support both
version in both commands or choosing one using it everywhere?
--
Jaime Casanova
Robert Haas robertmh...@gmail.com wrote:
I'm not sure why we're so glibly rejecting Dan's original
proposal. Sure, adjusting pg_upgrade when we whack around
pg_statistic is work, but who ever said that a workable in-place
upgrade facility would be maintenance-free? We're operating under
a
That's a speedup of nearly a factor of two, so clearly fsync-related
stalls are a big problem here, even with wal_buffers cranked up
through the ceiling.
H. Do you have any ability to test on XFS?
--
Josh Berkus
PostgreSQL Experts Inc.
http://pgexperts.com
--
Sent via pgsql-hackers
Jaime Casanova ja...@2ndquadrant.com wrote:
Has anyone ever complained about the inconsistency (at least it
seems that to me) of using INHERITS in CREATE TABLE and INHERIT in
ALTER TABLE?
They make sense to me as an English speaker. CREATE TABLE more or
less has a *description* of the
On Tue, Mar 13, 2012 at 3:30 PM, Robert Haas robertmh...@gmail.com wrote:
On Tue, Mar 13, 2012 at 5:42 PM, Bruce Momjian br...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run vacuumdb twice, once
with
On 03/13/2012 06:30 PM, Robert Haas wrote:
On Tue, Mar 13, 2012 at 5:42 PM, Bruce Momjianbr...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run vacuumdb twice, once
with default_statistics_target = 4, and
All,
I've discovered a built-in performance issue with replication failover
at one site, which I couldn't find searching the archives. I don't
really see what we can do to fix it, so I'm posting it here in case
others might have clever ideas.
1. The Free Space Map is not replicated between
On Tue, Mar 13, 2012 at 05:33:29PM -0500, Kevin Grittner wrote:
Bruce Momjian br...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run vacuumdb
twice, once with default_statistics_target = 4, and another
Bruce Momjian br...@momjian.us writes:
Another idea is to just copy over pg_statistic like we copy of
pg_largeobject now, and force autovacuum to run.
That would be an automatic crash in a 9.1 to 9.2 migration, as well as
any other release where we changed the column layout of pg_statistic.
On Tue, Mar 13, 2012 at 08:30:17PM -0400, Tom Lane wrote:
Bruce Momjian br...@momjian.us writes:
Another idea is to just copy over pg_statistic like we copy of
pg_largeobject now, and force autovacuum to run.
That would be an automatic crash in a 9.1 to 9.2 migration, as well as
any other
On Tue, Mar 13, 2012 at 08:22:51PM -0400, Bruce Momjian wrote:
On Tue, Mar 13, 2012 at 05:33:29PM -0500, Kevin Grittner wrote:
Bruce Momjian br...@momjian.us wrote:
What is the target=10 duration? I think 10 is as low as we can
acceptably recommend. Should we recommend they run
On Tue, Mar 13, 2012 at 4:53 PM, Josh Berkus j...@agliodbs.com wrote:
All,
I've discovered a built-in performance issue with replication failover
at one site, which I couldn't find searching the archives. I don't
really see what we can do to fix it, so I'm posting it here in case
others
On Wed, Mar 14, 2012 at 7:20 AM, Robert Haas robertmh...@gmail.com wrote:
On Tue, Mar 13, 2012 at 3:48 PM, Robert Haas robertmh...@gmail.com wrote:
On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes jeff.ja...@gmail.com wrote:
Rerunning all 4 benchmarks (both 16MB and 32MB wal_buffers on both
On Wed, Mar 14, 2012 at 8:53 AM, Josh Berkus j...@agliodbs.com wrote:
All,
I've discovered a built-in performance issue with replication failover
at one site, which I couldn't find searching the archives. I don't
really see what we can do to fix it, so I'm posting it here in case
others
On Tue, Mar 13, 2012 at 10:02 PM, Fujii Masao masao.fu...@gmail.com wrote:
On Wed, Mar 14, 2012 at 7:20 AM, Robert Haas robertmh...@gmail.com wrote:
On Tue, Mar 13, 2012 at 3:48 PM, Robert Haas robertmh...@gmail.com wrote:
On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes jeff.ja...@gmail.com wrote:
I have created a 29GB test database by using standard pgbnech -i -s
2000. Then I executed:
explain (analyze, buffers) select * from pgbench_accounts where aid in
(select cast(random()*2 as int) from generate_series(1,500));
Nested Loop (cost=30.00..6075.07 rows=1 width=97)
On Tue, Mar 13, 2012 at 7:05 PM, Fujii Masao masao.fu...@gmail.com wrote:
If it's really a high-UPDATE workload, wouldn't autovacuum start soon?
Also, while vacuum cleanup records are applied, could not the standby
also update its free space map, without having to send the actual FSM
updates? I
On Tue, Mar 13, 2012 at 6:44 PM, Josh Berkus j...@agliodbs.com wrote:
That's a speedup of nearly a factor of two, so clearly fsync-related
stalls are a big problem here, even with wal_buffers cranked up
through the ceiling.
H. Do you have any ability to test on XFS?
It seems I do.
On Tue, Mar 13, 2012 at 5:06 PM, Andres Freund and...@anarazel.de wrote:
Generally, uppon rereading, I have to say that I am not very happy with the
decision that ANY triggers are fired from other places than the specific
triggers. That seams to be a rather dangerous/confusing route to me.
I
On Tue, Mar 13, 2012 at 01:46:24PM -0400, Robert Haas wrote:
On Mon, Mar 12, 2012 at 3:28 PM, Simon Riggs si...@2ndquadrant.com wrote:
I agree with you that some worst case performance tests should be
done. Could you please say what you think the worst cases would be, so
those can be
On 13-03-2012 21:34, Bruce Momjian wrote:
It might be a solution for cases where we don't modify it. I frankly am
worried that if we copy over statistics even in ASCII that don't match
what the server expects, it might lead to a crash, which has me back to
wanting to speed up vacuumdb.
That
Patch attached.
Now I feel much better about it. Most people will either have fadvise, a
write cache (rightly or wrongly), or actually need the sync. Those that
have none of those can use -N.
Regards,
Jeff Davis
initdb-fsync-20120313.patch.gz
Description: GNU Zip compressed data
--
Sent
On Tue, Mar 13, 2012 at 1:53 PM, Robert Haas robertmh...@gmail.com wrote:
What I mean is that the function ArchiveEntry() is defined in
pg_backup_archiver.c, and it takes an argument called relpages, and
the string relpages does not appear anywhere else in that file.
Uhm, that's kinda
92 matches
Mail list logo