On Sun, Jul 25, 2010 at 4:32 PM, Robert Haas wrote:
> On Sun, Jul 25, 2010 at 2:37 AM, Jaime Casanova wrote:
>> but i have a few questions, maybe is right what you did i only want to
>> understand it:
>> - you added this in include/storage/smgr.h, so why is safe to assume
>> that if the backend !
On Jul29, 2010, at 00:48 , Greg Smith wrote:
> Finally got around to taking a longer look at your patch, sorry about the
> delay here. Patch itself seems to work on simple tests anyway (more on the
> one suspect bit below). You didn't show what the output looks like, so let's
> start with that b
On Wed, 2010-07-28 at 20:25 -0400, Tom Lane wrote:
> string_to_array() seems fine to me: it's a predictable transformation
> from text to text. However, I think that there really is an issue with
> array_to_string(), because that takes an anyarray and invokes the array
> element's type output func
Jeff Davis writes:
> On Tue, 2010-07-20 at 11:31 +0200, Pavel Stehule wrote:
>> I am working on to_array, to_string functions and I am looking on
>> string_to_array function. I am surprised so this function is marked as
>> immutable
> What's wrong with that? "current_date" is the part that's chan
Robert Haas writes:
> So now we have the following issues remaining:
> * page corruption after moving tablespace
> * ExplainOnePlan handles snapshots differently than ProcessQuery
> * name and comment of XLogSetAsyncCommitLSN() should be changed
> * Documentation fails to build as PDF
> ...and I
Oleg Bartunov writes:
> you can download dump http://mira.sai.msu.su/~megera/tmp/search_tab.dump
Hmm ... I'm not sure why you're failing to reproduce it, because it's
falling over pretty easily for me. After poking at it for awhile,
I am of the opinion that scanGetItem's handling of multiple key
On Wed, Jul 21, 2010 at 9:22 AM, Robert Haas wrote:
> On the other hand, if you have technical corrections, or if
> you have suggestions on how to do the same things better (rather than
> suggestions on what to do differently), that would be greatly
> appreciated.
Somewhere in that wiki page ther
On Tue, 2010-07-20 at 11:31 +0200, Pavel Stehule wrote:
> Hello
>
> I am working on to_array, to_string functions and I am looking on
> string_to_array function. I am surprised so this function is marked as
> immutable
>
> postgres=# select array_to_string(array[current_date],',');
> array_to_st
On Tue, Jul 27, 2010 at 4:09 PM, Tom Lane wrote:
> Robert Haas writes:
>> Well, that's pretty much saying we won't release before September.
>
> Yup, that's what I think. In fact I think September might be
> optimistic. This is what happens when you fork early and allow
> developers to start fo
Finally got around to taking a longer look at your patch, sorry about
the delay here. Patch itself seems to work on simple tests anyway (more
on the one suspect bit below). You didn't show what the output looks
like, so let's start with that because it is both kind of neat and not
what I expect
On Wed, 2010-07-28 at 15:37 -0400, Tom Lane wrote:
> So nevermind that distraction. I'm back to thinking that fix1 is
> the way to go.
Agreed.
It's uncontroversial to have a simple guard against corrupting an
uninitialized page, and uncontroversial is good for things that will be
back-patched.
On Jul 28, 2010, at 9:53 AM, Kris Jurka wrote:
> Technically you won't get NotificationResponse until transaction end, so you
> don't need to worry about that mid copy.
Ah, thanks for noting that. It would appear my original reading of the async
section didn't get far enough beyond "Frontends mu
On Wed, Jul 28, 2010 at 3:16 PM, Tom Lane wrote:
> (1) You're assuming that the page will be zeroes on the slave without
> having forced it to be so. A really obvious case where this fails
> is where we're doing crash-and-restart on the master: a later action
> could have modified the page away f
I wrote:
>>> I think it is appropriate to be setting the LSN/TLI in the case of a
>>> page that's been constructed by the caller as part of the WAL-logged
>>> action, but doing so in copy_relation_data seems rather questionable.
BTW, I thought of an argument that explains why that's sane: it marks
Robert Haas writes:
> On Wed, Jul 28, 2010 at 2:21 PM, Tom Lane wrote:
>> I've caught up on the thread now, and I think that fix2 (skip logging
>> the page) is extremely dangerous and has little if anything in its
>> favor.
> Why do you think that? They will be different only in terms of
> whet
On Wed, Jul 28, 2010 at 3:08 PM, Jeff Davis wrote:
> On Wed, 2010-07-28 at 14:50 -0400, Robert Haas wrote:
>> It seems like if log_newpage() were to set the LSN/TLI before calling
>> XLogInsert() - or optionally not - then it wouldn't be necessary to
>> set them also in heap_xlog_newpage(); the me
On Wed, 2010-07-28 at 14:50 -0400, Robert Haas wrote:
> It seems like if log_newpage() were to set the LSN/TLI before calling
> XLogInsert() - or optionally not - then it wouldn't be necessary to
> set them also in heap_xlog_newpage(); the memcpy operation would by
> definition have copied the righ
[ gradually catching up on email ]
Robert Haas writes:
> On Fri, Jul 16, 2010 at 2:39 PM, Tom Lane wrote:
>> I don't like the way you did that either (specifically, not the kluge
>> in NUMERIC_DIGITS()). It would probably work better if you declared
>> two different structs, or a union of same,
On Wed, Jul 28, 2010 at 2:21 PM, Tom Lane wrote:
> Robert Haas writes:
>> On Wed, Jul 28, 2010 at 12:36 PM, Tom Lane wrote:
>>> I understand it, and I don't like it one bit. I haven't caught up on
>>> this thread yet, but I think the only acceptable solution is one that
>>> leaves the slave in
Robert Haas writes:
> On Wed, Jul 28, 2010 at 12:36 PM, Tom Lane wrote:
>> I understand it, and I don't like it one bit. I haven't caught up on
>> this thread yet, but I think the only acceptable solution is one that
>> leaves the slave in the *same* state as the master.
> I might be missing so
On Wed, 2010-07-28 at 13:18 -0400, Robert Haas wrote:
> In Jeff's original example, he crashes the database
> after extending the relation but before initializing and writing the
> new page. I believe that at that point no XLOG has been written yet,
> so the relation has been extended but there is
On Wed, Jul 28, 2010 at 12:36 PM, Tom Lane wrote:
> Jeff Davis writes:
>> However, when Simon said "We definitely shouldn't do anything that
>> leaves standby different to primary." you said "obviously". Fix2 can
>> leave a difference between the two, because zeroed pages at the end of
>> the hea
On Wed, 2010-07-28 at 12:36 -0400, Tom Lane wrote:
> Jeff Davis writes:
> > However, when Simon said "We definitely shouldn't do anything that
> > leaves standby different to primary." you said "obviously". Fix2 can
> > leave a difference between the two, because zeroed pages at the end of
> > the
On Wed, 28 Jul 2010, James William Pye wrote:
hrm, I suppose a lazy way around that problem would be to suspend all
client messages(client_min_messages) during COPY IN. Tho, I guess one
would still have to contend with NotificationResponse, and
ParameterStatus..
Technically you won't get
Jeff Davis writes:
> However, when Simon said "We definitely shouldn't do anything that
> leaves standby different to primary." you said "obviously". Fix2 can
> leave a difference between the two, because zeroed pages at the end of
> the heap file on the primary will not be sent to the standby (th
On Wed, 2010-07-28 at 09:30 -0700, David E. Wheeler wrote:
> On Jul 28, 2010, at 7:57 AM, Joshua D. Drake wrote:
>
> > Hah! I gave you kudos :P (you are in the FAQ)
>
> Ah, thanks. The link is missing a "G": It's "PGXN," not "PXN".
Yeah that is already fixed, just waiting for cache to clear (on
On Jul 25, 2010, at 8:01 AM, Kris Jurka wrote:
> The JDBC driver reads server messages for multiple reasons.
> One of them is indeed to do early failure detection.
That's high quality. =)
> Another is to pickup NoticeResponse messages to avoid a network buffer
> deadlock.
That's a good catch.
On Jul 28, 2010, at 7:57 AM, Joshua D. Drake wrote:
> Hah! I gave you kudos :P (you are in the FAQ)
Ah, thanks. The link is missing a "G": It's "PGXN," not "PXN".
Best,
David
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www
On Wed, 2010-07-28 at 06:40 -0400, Robert Haas wrote:
> > fix1 -- Only call PageSetLSN/TLI inside log_newpage() and
> > heap_xlog_newpage() if the page is not zeroed.
> >
> > fix2 -- Don't call log_newpage() at all if the page is not zeroed.
> >
> > Please review. I don't have a strong opinion abou
On Wed, Jul 28, 2010 at 4:54 PM, Bruce Momjian wrote:
> Are we basically leaving pgAdmin in this state until we come up with a
> fix and need a new minor release? We pride ourselves in not introducing
> breakage in minor releases, but it has certainly happened in this case,
> and it is making pgA
On Wed, 2010-07-28 at 15:24 +0300, Peter Eisentraut wrote:
> On tor, 2010-07-15 at 10:24 +0100, Simon Riggs wrote:
> > Patch to reduce lock levels for
> > ALTER TABLE
> > CREATE TRIGGER
> > CREATE RULE
>
> Tried this out, but $subject is still the case. The problem is that
> ATRewriteCatalogs
On Tue, 27 Jul 2010 19:55:18 -0700, "David E. Wheeler"
wrote:
> On Jul 27, 2010, at 3:01 PM, Joshua D. Drake wrote:
>
>> Correct. We are also hoping to get some sponsorship for it.
>>
>> https://www.fossexperts.com/
>
> Frigging copycat.
Hah! I gave you kudos :P (you are in the FAQ)
JD
--
P
Yeb Havinga writes:
> Robert Haas wrote:
>> On Wed, Jul 28, 2010 at 7:24 AM, Yeb Havinga wrote:
>>> Wouldn't it be relatively easy, to rewrite the filter expression by adding
>>> expressions, instead of replacing constants, in the disjunctive case, so the
>>> example at hand would become:
>>>
>>
On Wed, 2010-07-28 at 15:24 +0300, Peter Eisentraut wrote:
> On tor, 2010-07-15 at 10:24 +0100, Simon Riggs wrote:
> > Patch to reduce lock levels for
> > ALTER TABLE
> > CREATE TRIGGER
> > CREATE RULE
>
> Tried this out, but $subject is still the case. The problem is that
> ATRewriteCatalogs
On tor, 2010-07-15 at 10:24 +0100, Simon Riggs wrote:
> Patch to reduce lock levels for
> ALTER TABLE
> CREATE TRIGGER
> CREATE RULE
Tried this out, but $subject is still the case. The problem is that
ATRewriteCatalogs() calls AlterTableCreateToastTable() based on what it
thinks the subcomman
Robert Haas wrote:
On Wed, Jul 28, 2010 at 7:24 AM, Yeb Havinga wrote:
Sorry? I though what Equivalence Class provides is the "proving" that
using this qualification or another will *not* affect the output.
In a query like...
SELECT d1.ID, d2.ID
FROM DocPrimary d1
JOIN DocPrim
Hello,
can you send a current version, please. I looked to git repository,
but you did more changes.
Thank you
Pavel Stehule
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
Robert Haas writes:
> SELECT d1.ID, d2.ID
> FROM DocPrimary d1
>JOIN DocPrimary d2 ON d2.BasedOn=d1.ID
> WHERE (d1.ID=234409763) or (d2.ID=234409763)
>
> ...you're going to scan d1, scan d2, and then join the results. The
> scan of d1 is going to produce different results depending on whet
On Wed, Jul 28, 2010 at 7:24 AM, Yeb Havinga wrote:
>>> Sorry? I though what Equivalence Class provides is the "proving" that
>>> using this qualification or another will *not* affect the output.
>>
>> In a query like...
>>
>> SELECT d1.ID, d2.ID
>> FROM DocPrimary d1
>> JOIN DocPrimary d2 ON
On Wed, Jul 28, 2010 at 7:02 AM, Simon Riggs wrote:
> On Tue, 2010-07-27 at 21:23 -0700, Jeff Davis wrote:
>
>> Both potential fixes attached and both appear to work.
>>
>> fix1 -- Only call PageSetLSN/TLI inside log_newpage() and
>> heap_xlog_newpage() if the page is not zeroed.
>>
>> fix2 -- Don
Andrew Dunstan wrote:
No, the configure test is wrong. Here's what's in configure.ac:
dnl Temporary hack until minimum PostgreSQL version is 8.5:
dnl If PostgreSQL < 8.5 is detected, trigger the inclusion of the
new versioned PGXS targets
PGXSOVERRIDE=0
if test ! "$PGSQL_MINOR_V
Robert Haas wrote:
On Wed, Jul 28, 2010 at 6:55 AM, Dimitri Fontaine
wrote:
Robert Haas writes:
But here you want to have different paths for
the same relation that generate *different output*, and the planner
doesn't understand that concept.
Sorry? I though what Equivalence
On Wed, Jul 28, 2010 at 6:55 AM, Dimitri Fontaine
wrote:
> Robert Haas writes:
>> But here you want to have different paths for
>> the same relation that generate *different output*, and the planner
>> doesn't understand that concept.
>
> Sorry? I though what Equivalence Class provides is the "p
Mark Cave-Ayland wrote:
Andrew Dunstan wrote:
The real problem has nothing to do with any of the analysis, as you
say. It is this: they have an override file for PGXS and it uses
$(mkinstalldirs) which we got rid of about a year ago. So apparently
they haven't been testing much against any
On Tue, 2010-07-27 at 21:23 -0700, Jeff Davis wrote:
> Both potential fixes attached and both appear to work.
>
> fix1 -- Only call PageSetLSN/TLI inside log_newpage() and
> heap_xlog_newpage() if the page is not zeroed.
>
> fix2 -- Don't call log_newpage() at all if the page is not zeroed.
>
>
Robert Haas writes:
> But here you want to have different paths for
> the same relation that generate *different output*, and the planner
> doesn't understand that concept.
Sorry? I though what Equivalence Class provides is the "proving" that
using this qualification or another will *not* affect
On Wed, Jul 28, 2010 at 3:45 AM, Dimitri Fontaine
wrote:
>> Even if we understood how to direct the rewriting process, I'm really
>> dubious that it would win often enough to justify the added planning
>> time. The particular problem here seems narrow enough that solving it
>> on the client side
On Wed, Jul 28, 2010 at 12:23 AM, Jeff Davis wrote:
> On Tue, 2010-07-27 at 15:23 -0700, Jeff Davis wrote:
>> On Tue, 2010-07-27 at 17:18 -0400, Robert Haas wrote:
>> > > My first concern with that idea was that it may create an inconsistency
>> > > between the primary and the standby. The primary
Andrew Dunstan wrote:
The real problem has nothing to do with any of the analysis, as you say.
It is this: they have an override file for PGXS and it uses
$(mkinstalldirs) which we got rid of about a year ago. So apparently
they haven't been testing much against any of our alphas or betas or
On Wed, Jul 28, 2010 at 6:08 AM, Boxuan Zhai wrote:
>> On Tue, Jul 27, 2010 at 1:04 AM, Boxuan Zhai wrote:
>> > I have get a edition that the merge command can run. It accept the
>> > standard
>> > merge command and can do UPDATE, INSERT and DELETE actions now. But we
>> > cannot put additional q
2010/7/28 Robert Haas
> On Tue, Jul 27, 2010 at 1:04 AM, Boxuan Zhai wrote:
> > I have get a edition that the merge command can run. It accept the
> standard
> > merge command and can do UPDATE, INSERT and DELETE actions now. But we
> > cannot put additional qualification for actions. There are
Tom Lane writes:
> In the example, we do have d1.id and d2.basedon grouped in an
> equivalence class. So in principle you could substitute d1.id into the
> WHERE clause in place of d2.basedon, once you'd checked that it was
> being used with an operator that's compatible with the specific
> equiv
52 matches
Mail list logo