On Mon, Jun 18, 2012 at 09:30:14PM -0400, Robert Haas wrote:
> There might be something to the idea of demoting a few of the things
> we've traditionally had as NOTICEs, though. IME, the following two
> messages account for a huge percentage of the chatter:
>
> NOTICE: CREATE TABLE will create i
On tis, 2012-06-19 at 02:38 -0400, Tom Lane wrote:
> Peter Eisentraut writes:
> > On mån, 2012-06-18 at 17:57 -0400, James Cloos wrote:
> >> I presume that lc_ctype is the significant difference?
>
> > It certainly makes some difference, but it's a bit shocking that
> makes
> > things that much s
On Tuesday, June 19, 2012 12:47:54 AM Christopher Browne wrote:
> On Mon, Jun 18, 2012 at 11:50 AM, Andres Freund
wrote:
> > Hi Simon,
> >
> > On Monday, June 18, 2012 05:35:40 PM Simon Riggs wrote:
> >> On 13 June 2012 19:28, Andres Freund wrote:
> >> > This adds a new configuration parameter
On tor, 2012-06-14 at 13:38 -0400, Robert Haas wrote:
> On Sun, Jan 8, 2012 at 3:48 PM, Peter Eisentraut wrote:
> > psql tab completion currently only supports the form GRANT privilege ON
> > something TO someone (and the analogous REVOKE), but not the form GRANT
> > role TO someone. Here is a pa
Hi,
On Tuesday, June 19, 2012 08:03:04 AM Tom Lane wrote:
> Andres Freund writes:
> > On Monday, June 18, 2012 11:51:27 PM Daniel Farina wrote:
> >> What's the cost of going a lot higher? Because if one makes enough
> >> numerical space available, one can assign node identities without a
> >> co
On Tuesday, June 19, 2012 04:12:47 AM Steve Singer wrote:
> On 12-06-18 07:30 AM, Andres Freund wrote:
> > Hrmpf #666. I will go through through the series commit-by-commit again
> > to make sure everything compiles again. Reordinging this late definitely
> > wasn't a good idea...
> >
> > I pushed
Thank you.
> What happens if the server skips an end-of-recovery checkpoint,
> is promoted to the master, runs some write transactions,
> crashes and restarts automatically before it completes
> checkpoint? In this case, the server needs to do crash recovery
> from the last checkpoint record with
There might be something to the idea of demoting a few of the things
we've traditionally had as NOTICEs, though. IME, the following two
messages account for a huge percentage of the chatter:
NOTICE: CREATE TABLE will create implicit sequence "foo_a_seq" for
serial column "foo.a"
NOTICE: CREA
On Tue, Jun 19, 2012 at 4:43 AM, Robert Haas wrote:
>
>
> Please add this patch here so that it doesn't get lost in the shuffle:
>
> https://commitfest.postgresql.org/action/commitfest_view/open
Hmm, that raises an interesting question (though maybe I've just
missed this happening in the past). T
2012-06-18 19:46 keltezéssel, Alvaro Herrera írta:
Excerpts from Boszormenyi Zoltan's message of vie may 11 03:54:13 -0400 2012:
Hi,
another rebasing and applied the GIT changes in
ada8fa08fc6cf5f199b6df935b4d0a730aaa4fec to the
Windows implementation of PGSemaphoreTimedLock.
Hi,
I gave the f
> I'm almost inclined to suggest that we not get next-LSN from WAL, but
> by scanning all the pages in the main data store and computing the max
> observed LSN. This is clearly not very attractive from a performance
> standpoint, but it would avoid the obvious failure mode where you lost
> some re
Hi,
> -Original Message-
> From: Robert Haas [mailto:robertmh...@gmail.com]
> Sent: Tuesday, June 19, 2012 3:12 AM
> To: Ants Aasma
> Cc: Etsuro Fujita; Jay Levitt; Tom Lane; PostgreSQL-development; Francois
> Deliege
> Subject: Re: [HACKERS] [PATCH] Lazy hashaggregate when no aggregation
Dave Page writes:
> On Tue, Jun 19, 2012 at 4:43 AM, Robert Haas wrote:
>>
>>
>> Please add this patch here so that it doesn't get lost in the shuffle:
>>
>> https://commitfest.postgresql.org/action/commitfest_view/open
>
> Hmm, that raises an interesting question (though maybe I've just
> miss
On Tue, Jun 19, 2012 at 11:04 AM, Alex Shulgin wrote:
>
> Dave Page writes:
>
>> On Tue, Jun 19, 2012 at 4:43 AM, Robert Haas wrote:
>>>
>>>
>>> Please add this patch here so that it doesn't get lost in the shuffle:
>>>
>>> https://commitfest.postgresql.org/action/commitfest_view/open
>>
>> Hmm,
Dave Page writes:
> On Tue, Jun 19, 2012 at 11:04 AM, Alex Shulgin wrote:
>>
>> In a real bug-tracking system we would create a new bug/ticket and set
>> it's target version to 'candidate for next minor release' or something
>> like that. This way, if we don't release unless all targeted bugs
On ons, 2012-01-18 at 21:21 +0200, Peter Eisentraut wrote:
> On lör, 2012-01-07 at 16:41 -0500, Tom Lane wrote:
> > Peter Eisentraut writes:
> > > I suggest that we change PostgresMain(), PostmasterMain(), BackendRun(),
> > > WalSenderMain(), and WalSndLoop() to return void as well.
> >
> > I agr
2012/6/18 Merlin Moncure :
> I can't help but wonder (having been down the contrib/core/extension
> road myself) if it isn't better to improve the facilities to register
> and search for qualified extensions (like Perl CPAN) so that people
> looking for code to improve their backends can find it.
What is the latest theory on using int4 vs. int32 in C code?
(equivalently int2, int16)
I had the idea that using int4 was sort of deprecated, and most code
uses int32, but I've come across several uses of int4 lately that looked
odd to me.
I think the main reason that we define int4 in C is for
Peter Eisentraut writes:
> On tis, 2012-06-19 at 02:38 -0400, Tom Lane wrote:
>> If James is testing text-comparison-heavy operations, it doesn't seem
>> shocking in the least. strcoll() in most non-C locales is a pig.
> Ah yes, of course, having lc_ctype != C also selects strcoll instead of
>
Peter Eisentraut writes:
> What is the latest theory on using int4 vs. int32 in C code?
> (equivalently int2, int16)
I thought the general idea was to use int32 most places, but int4 in
catalog declarations. I don't think it's tremendously important if
somebody uses the other though.
> While we
On Mon, Jun 18, 2012 at 11:01 PM, Robert Haas wrote:
> On Mon, Jun 18, 2012 at 12:24 PM, Merlin Moncure wrote:
>> I can't help but wonder (having been down the contrib/core/extension
>> road myself) if it isn't better to improve the facilities to register
>> and search for qualified extensions (l
On Tue, Jun 19, 2012 at 2:44 AM, Amit Kapila wrote:
>> AFAIR you can create pg_control from scratch already with pg_resetxlog.
>> The hard part is coming up with values for the counters, such as the
>> next WAL location. Some of them such as next OID are pretty harmless
>> if you don't guess righ
Andres Freund writes:
> On Tuesday, June 19, 2012 08:03:04 AM Tom Lane wrote:
>> "Every WAL record"? Why in heck would you attach it to every record?
>> Surely putting it in WAL page headers would be sufficient.
> The idea is that you can have cascading, circular and whatever replication
> topo
Hi,
There are 70+ calls of malloc in the backend in the form of
type* foo = malloc(sizeof(...));
if(!foo)
elog(ERROR, "could not allocate memory");
which is a bit annoying to write at times. Would somebody argue against
introducing a function that does the above named xmalloc() or malloc_or_
On Tuesday, June 19, 2012 04:17:01 PM Tom Lane wrote:
> Andres Freund writes:
> > On Tuesday, June 19, 2012 08:03:04 AM Tom Lane wrote:
> >> "Every WAL record"? Why in heck would you attach it to every record?
> >> Surely putting it in WAL page headers would be sufficient.
> >
> > The idea is th
Andres Freund writes:
> On Tuesday, June 19, 2012 04:17:01 PM Tom Lane wrote:
>> ... (If you are thinking
>> of something sufficiently high-level that merging could possibly work,
>> then it's not WAL, and we shouldn't be trying to make the WAL
>> representation cater for it.)
> The idea is that
Andres Freund writes:
> There are 70+ calls of malloc in the backend in the form of
> type* foo = malloc(sizeof(...));
> if(!foo)
>elog(ERROR, "could not allocate memory");
> which is a bit annoying to write at times. Would somebody argue against
> introducing a function that does the above
So, just to give a bit more weight to my argument that we should
recognise that equivalent strings ought to be treated identically, I
direct your attention to conformance requirement C9 of Unicode 3.0:
http://www.unicode.org/unicode/standard/versions/enumeratedversions.html#Unicode_3_0_0
This "re
On Tuesday, June 19, 2012 04:30:59 PM Tom Lane wrote:
> Andres Freund writes:
> > On Tuesday, June 19, 2012 04:17:01 PM Tom Lane wrote:
> >> ... (If you are thinking
> >> of something sufficiently high-level that merging could possibly work,
> >> then it's not WAL, and we shouldn't be trying to m
On Tuesday, June 19, 2012 04:38:56 PM Tom Lane wrote:
> Andres Freund writes:
> > There are 70+ calls of malloc in the backend in the form of
> >
> > type* foo = malloc(sizeof(...));
> > if(!foo)
> >
> >elog(ERROR, "could not allocate memory");
> >
> > which is a bit annoying to write at ti
Excerpts from Boszormenyi Zoltan's message of mar jun 19 04:44:35 -0400 2012:
>
> 2012-06-18 19:46 keltezéssel, Alvaro Herrera írta:
> > Excerpts from Boszormenyi Zoltan's message of vie may 11 03:54:13 -0400
> > 2012:
> >> Hi,
> >>
> >> another rebasing and applied the GIT changes in
> >> ada8f
2012/6/19 Merlin Moncure :
> On Mon, Jun 18, 2012 at 11:01 PM, Robert Haas wrote:
>> On Mon, Jun 18, 2012 at 12:24 PM, Merlin Moncure wrote:
>>> I can't help but wonder (having been down the contrib/core/extension
>>> road myself) if it isn't better to improve the facilities to register
>>> and s
Peter Geoghegan wrote:
> So, just to give a bit more weight to my argument that we should
> recognise that equivalent strings ought to be treated identically
Since we appear to be questioning everything in this area, I'll
raise something which has been bugging me for a while: in some other
sys
On Mon, Jun 18, 2012 at 09:34:30PM +0300, Peter Eisentraut wrote:
> On mån, 2012-06-18 at 18:05 +0200, Andres Freund wrote:
> > - defaulting to initdb -N in the regression suite is not a good imo,
> > because that way the buildfarm won't catch problems in that area...
> >
> The regression test sui
On Mon, Jun 18, 2012 at 1:42 PM, Martijn van Oosterhout
wrote:
> On Sun, Jun 17, 2012 at 12:29:53PM -0400, Tom Lane wrote:
>> The fly in the ointment with any of these ideas is that the "configure
>> list" is not a list of exact cipher names, as per Magnus' comment that
>> the current default incl
On Mon, Jun 18, 2012 at 3:30 PM, Alvaro Herrera
wrote:
> Excerpts from Alex Hunsaker's message of vie feb 10 16:53:05 -0300 2012:
>
>> Seems like we missed the fact that we still did SvUTF8_on() in sv2cstr
>> and SvPVUTF8() when turning a perl string into a cstring.
>
> Hmm, this patch belongs int
On Tuesday, June 19, 2012 10:14:08 AM Heikki Linnakangas wrote:
> On 18.06.2012 21:08, Heikki Linnakangas wrote:
> > On 18.06.2012 21:00, Robert Haas wrote:
> >> On Thu, Jun 14, 2012 at 5:58 PM, Andres Freund
> >>
> >> wrote:
> 1. Use a 64-bit segment number, instead of the log/seg combinatio
On Tue, Jun 19, 2012 at 2:15 AM, Tom Lane wrote:
> Robert Haas writes:
>> There might be something to the idea of demoting a few of the things
>> we've traditionally had as NOTICEs, though. IME, the following two
>> messages account for a huge percentage of the chatter:
>
>> NOTICE: CREATE TABL
On Tue, Jun 19, 2012 at 4:14 AM, Heikki Linnakangas
wrote:
> Well, that was easier than I thought. Attached is a patch to make XLogRecPtr
> a uint64, on top of my other WAL format patches. I think we should go ahead
> with this.
+1.
> The LSNs on pages are still stored in the old format, to avoi
On Tue, Jun 19, 2012 at 5:30 PM, Kyotaro HORIGUCHI
wrote:
> Thank you.
>
>> What happens if the server skips an end-of-recovery checkpoint,
>> is promoted to the master, runs some write transactions,
>> crashes and restarts automatically before it completes
>> checkpoint? In this case, the server
On Tue, Jun 19, 2012 at 1:56 AM, Tom Lane wrote:
> The transaction would be committed before a command success report is
> delivered to the client, so I don't think delivered-and-not-marked is
> possible.
...unless you have configured synchronous_commit=off, or fsync=off.
Or unless your disk mel
Andres Freund writes:
> On Tuesday, June 19, 2012 04:30:59 PM Tom Lane wrote:
>>> ... (If you are thinking
>>> of something sufficiently high-level that merging could possibly work,
>>> then it's not WAL, and we shouldn't be trying to make the WAL
>>> representation cater for it.)
> Do you reall
On Tue, Jun 19, 2012 at 10:15 AM, Kohei KaiGai wrote:
> Let me push the pgsql_fdw in core from different perspective.
>
> Right now, FDW is a feature that will take many enhancement in
> the near future like join-pushdown, writable APIs and so on.
> If we would not have a FDW extension in core tha
On 19 June 2012 16:17, Kevin Grittner wrote:
> Peter Geoghegan wrote:
>
>> So, just to give a bit more weight to my argument that we should
>> recognise that equivalent strings ought to be treated identically
>
> Since we appear to be questioning everything in this area, I'll
> raise something wh
Hi Fujita-san,
Could you rebase this patch towards the latest tree?
It was unavailable to apply the patch cleanly.
I looked over the patch, then noticed a few points.
At ProcessCopyOptions(), defel->arg can be NIL, isn't it?
If so, cstate->convert_binary is not a suitable flag to check
redundant
Hi,
On Tuesday, June 19, 2012 06:11:20 PM Tom Lane wrote:
> Andres Freund writes:
> > On Tuesday, June 19, 2012 04:30:59 PM Tom Lane wrote:
> >>> ... (If you are thinking
> >>> of something sufficiently high-level that merging could possibly work,
> >>> then it's not WAL, and we shouldn't be try
Peter Geoghegan wrote:
> Kevin Grittner wrote:
>> Since we appear to be questioning everything in this area, I'll
>> raise something which has been bugging me for a while: in some
>> other systems I've used, the "tie-breaker" comparison for
>> equivalent values comes after equivalence sorting o
Our current interpretation of the difference between foreign keys with
ON UPDATE/DELETE NO ACTION and those with ON UPDATE/DELETE RESTRICT
is that they mean the same thing but RESTRICT checks are not deferrable.
It follows from this that the trigger code ought to be the same for
NO ACTION and RESTR
Excerpts from Boszormenyi Zoltan's message of mar jun 19 12:44:04 -0400 2012:
> OK, all 4 Check* functions are now moved back into proc.c,
> nothing outside of timeout.c touches anything in it. New patches
> are attached.
Yeah, I like this one better, thanks.
It seems to me that the "check" fun
On 19 June 2012 17:45, Kevin Grittner wrote:
> Peter Geoghegan wrote:
>> Are you sure that they actually have a tie-breaker, and don't just
>> make the distinction between equality and equivalence (if only
>> internally)?
>
> I'm pretty sure that when I was using Sybase ASE the order for
> non-eq
On Mon, 2012-06-18 at 21:41 +0200, Andres Freund wrote:
> It calls pg_flush_data inside of copy_file which does the posix_fadvise... So
> maybe just put the sync_file_range in pg_flush_data?
The functions in fd.c aren't linked to initdb, so it's a challenge to
share that code (I remember now: tha
On Tue, Jun 19, 2012 at 12:11 PM, Tom Lane wrote:
> Andres Freund writes:
>> On Tuesday, June 19, 2012 04:30:59 PM Tom Lane wrote:
... (If you are thinking
of something sufficiently high-level that merging could possibly work,
then it's not WAL, and we shouldn't be trying to make
On Tuesday, June 19, 2012 07:22:02 PM Jeff Davis wrote:
> On Mon, 2012-06-18 at 21:41 +0200, Andres Freund wrote:
> > It calls pg_flush_data inside of copy_file which does the
> > posix_fadvise... So maybe just put the sync_file_range in pg_flush_data?
>
> The functions in fd.c aren't linked to in
On Mon, Jun 18, 2012 at 1:30 PM, Alvaro Herrera
wrote:
>
> Excerpts from Alex Hunsaker's message of vie feb 10 16:53:05 -0300 2012:
>
>> Seems like we missed the fact that we still did SvUTF8_on() in sv2cstr
>> and SvPVUTF8() when turning a perl string into a cstring.
>
> Hmm, this patch belongs i
On Tue, Jun 19, 2012 at 10:17 AM, Andres Freund wrote:
> There are 70+ calls of malloc in the backend in the form of
>
> type* foo = malloc(sizeof(...));
> if(!foo)
> elog(ERROR, "could not allocate memory");
>
> which is a bit annoying to write at times. Would somebody argue against
> introduci
On Tuesday, June 19, 2012 07:35:53 PM Robert Haas wrote:
> On Tue, Jun 19, 2012 at 10:17 AM, Andres Freund
wrote:
> > There are 70+ calls of malloc in the backend in the form of
> >
> > type* foo = malloc(sizeof(...));
> > if(!foo)
> > elog(ERROR, "could not allocate memory");
> >
> > which i
Peter Geoghegan wrote:
> Kevin Grittner wrote:
>> I'm pretty sure that when I was using Sybase ASE the order for
>> non-equal values was always predictable, and it behaved in the
>> manner I describe below. I'm less sure about any other product.
>
> Maybe it used a physical row identifier as
Hi,
The most important part, even for people not following my discussion with
Robert is at the bottom where the possible wal decoding strategies are laid
out.
On Tuesday, June 19, 2012 03:20:58 AM Robert Haas wrote:
> On Sat, Jun 16, 2012 at 7:43 AM, Andres Freund
wrote:
> >> > Hm. Yes, you c
On 19 June 2012 18:57, Kevin Grittner wrote:
> We weren't using en_US.UTF-8 collation (or any other "proper"
> collation) on Sybase -- I'm not sure whether they even supported
> proper collation sequences on the versions we used. I'm thinking of
> when we were using their "case insensitive" sorti
On Fri, Jun 15, 2012 at 4:27 PM, Dimitri Fontaine
wrote:
> Allow me to open the new season of the DML trigger series, named
> pg_event_trigger. This first episode is all about setting up the drama,
> so that next ones make perfect sense.
Comments:
1. I still think we ought to get rid of the noti
"Kevin Grittner" writes:
> I wasn't aware that en_US.UTF-8 doesn't have equivalence without
> equality. I guess that surprising result in my last post is just
> plain inevitable with that collation then. Bummer. Is there
> actually anyone who finds that to be a useful behavior? For a
> collati
On 19 June 2012 19:44, Peter Geoghegan wrote:
> You could do that, and some people do use custom collations for
> various reasons. That's obviously very much of minority interest
> though. Most people will just use citext or something. However, since
> citext is itself a client of varstr_cmp(), th
Excerpts from Robert Haas's message of mar jun 19 11:36:41 -0400 2012:
>
> On Mon, Jun 18, 2012 at 3:30 PM, Alvaro Herrera
> wrote:
> > Excerpts from Alex Hunsaker's message of vie feb 10 16:53:05 -0300 2012:
> >
> >> Seems like we missed the fact that we still did SvUTF8_on() in sv2cstr
> >> an
On Tue, Jun 19, 2012 at 9:47 AM, Tom Lane wrote:
> Peter Eisentraut writes:
>> What is the latest theory on using int4 vs. int32 in C code?
>> (equivalently int2, int16)
>
> I thought the general idea was to use int32 most places, but int4 in
> catalog declarations. I don't think it's tremendous
On Wed, Jun 13, 2012 at 7:28 AM, Andres Freund wrote:
> Adds a single and a double linked list which can easily embedded into other
> datastructures and can be used without any additional allocations.
dllist.h advertises that it's embeddable. Can you use that instead,
or enhance it slightly to s
On Tuesday, June 19, 2012 07:24:13 PM Robert Haas wrote:
> On Tue, Jun 19, 2012 at 12:11 PM, Tom Lane wrote:
> > Andres Freund writes:
> >> On Tuesday, June 19, 2012 04:30:59 PM Tom Lane wrote:
> ... (If you are thinking
> of something sufficiently high-level that merging could possibl
On 19 June 2012 20:11, Robert Haas wrote:
> On Tue, Jun 19, 2012 at 9:47 AM, Tom Lane wrote:
>> Peter Eisentraut writes:
>>> What is the latest theory on using int4 vs. int32 in C code?
>>> (equivalently int2, int16)
>>
>> I thought the general idea was to use int32 most places, but int4 in
>> c
On Tue, Jun 19, 2012 at 1:43 AM, Tom Lane wrote:
> Amit Kapila writes:
>>> AFAIR you can create pg_control from scratch already with pg_resetxlog.
>>> The hard part is coming up with values for the counters, such as the
>>> next WAL location. Some of them such as next OID are pretty harmless
>>>
On Tue, Jun 19, 2012 at 3:03 PM, Alvaro Herrera
wrote:
> That's what I thought. I will commit to both branches soon, then.
I think there might be three branches involved.
> Mind you, this should have been an "open item", not a commitfest item.
> (Actually not even an open item. We should have
On Tue, Jun 19, 2012 at 7:31 AM, Peter Eisentraut wrote:
> On ons, 2012-01-18 at 21:21 +0200, Peter Eisentraut wrote:
>> On lör, 2012-01-07 at 16:41 -0500, Tom Lane wrote:
>> > Peter Eisentraut writes:
>> > > I suggest that we change PostgresMain(), PostmasterMain(), BackendRun(),
>> > > WalSende
On Tuesday, June 19, 2012 09:16:41 PM Robert Haas wrote:
> On Wed, Jun 13, 2012 at 7:28 AM, Andres Freund
wrote:
> > Adds a single and a double linked list which can easily embedded into
> > other datastructures and can be used without any additional allocations.
>
> dllist.h advertises that it'
On Wed, Jun 13, 2012 at 7:28 AM, Andres Freund wrote:
> From: Andres Freund
>
> The previous coding could miss xlog writeouts at several places. E.g. when wal
> was written out by the background writer or even after a commit if
> synchronous_commit=off.
> This could lead to delays in sending data
On Tue, Jun 19, 2012 at 3:48 PM, Andres Freund wrote:
> Why is that code not used more widely? Quite a bit of our list usage should be
> replaced embedding list element in larger structs imo. There are also open-
> coded inline list manipulations around (check aset.c for example).
Because we've g
On Wed, Jun 13, 2012 at 2:28 PM, Andres Freund wrote:
> +/*
> + * removes a node from a list
> + * Attention: O(n)
> + */
> +static inline void ilist_s_remove(ilist_s_head *head,
> + ilist_s_node *node)
> +{
> + ilist_s_node *last = &head->head;
> + ili
Hi,
On Tuesday, June 19, 2012 09:59:48 PM Marko Kreen wrote:
> On Wed, Jun 13, 2012 at 2:28 PM, Andres Freund
wrote:
> > +/*
> > + * removes a node from a list
> > + * Attention: O(n)
> > + */
> > +static inline void ilist_s_remove(ilist_s_head *head,
> > + ilist
On Tuesday, June 19, 2012 09:55:30 PM Robert Haas wrote:
> On Wed, Jun 13, 2012 at 7:28 AM, Andres Freund
wrote:
> > From: Andres Freund
> >
> > The previous coding could miss xlog writeouts at several places. E.g.
> > when wal was written out by the background writer or even after a commit
> >
On Tue, Jun 19, 2012 at 11:02 PM, Andres Freund wrote:
> On Tuesday, June 19, 2012 09:59:48 PM Marko Kreen wrote:
>> On Wed, Jun 13, 2012 at 2:28 PM, Andres Freund
> wrote:
>> > +/*
>> > + * removes a node from a list
>> > + * Attention: O(n)
>> > + */
>> > +static inline void ilist_s_remove(ilis
On Tuesday, June 19, 2012 09:58:43 PM Robert Haas wrote:
> On Tue, Jun 19, 2012 at 3:48 PM, Andres Freund
wrote:
> > Why is that code not used more widely? Quite a bit of our list usage
> > should be replaced embedding list element in larger structs imo. There
> > are also open- coded inline list
Andres Freund wrote:
> Robert Haas wrote:
>> Tom Lane wrote:
>>> However, if we're dead set on doing it that way, let us put
>>> information that is only relevant to logical replication records
>>> into only the logical replication records.
>> Right. If we decide we need this, and if we did de
On Mon, Jun 18, 2012 at 8:42 PM, Jeff Janes wrote:
> There was a regression introduced in 9.2 that effects the creation and
> loading of lots of small tables in a single transaction.
>
> It affects the loading of a pg_dump file which has a large number of
> small tables (10,000 schemas, one table
On Mon, Jun 18, 2012 at 6:35 PM, Simon Riggs wrote:
> On 13 June 2012 19:28, Andres Freund wrote:
>> This adds a new configuration parameter multimaster_node_id which determines
>> the id used for wal originating in one cluster.
>
> Looks good and it seems this aspect at least is commitable in th
On 19 June 2012 17:48, Tom Lane wrote:
> I think that the argument for having the RESTRICT triggers behave
> like this is that the SQL spec envisions the RESTRICT check occurring
> immediately when the individual PK row is updated/deleted, and so there
> would be no opportunity for another PK row
On tis, 2012-06-19 at 09:33 -0400, Tom Lane wrote:
> Come to think of it, another possible factor is that LIKE can't use
> ordinary indexes on text if the locale isn't C.
But he reported that the plans are the same.
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make c
Andres Freund wrote:
> The problem is just that to support basically arbitrary decoding
> requirements you need to provide at least those pieces of
> information in a transactionally consistent manner:
> * the data
> * table names
> * column names
> * type information
> * replication configurati
On Tue, Jun 19, 2012 at 4:33 PM, Robert Haas wrote:
> On Mon, Jun 18, 2012 at 8:42 PM, Jeff Janes wrote:
>> There was a regression introduced in 9.2 that effects the creation and
>> loading of lots of small tables in a single transaction.
>>
>> It affects the loading of a pg_dump file which has a
On 20 June 2012 04:58, Marko Kreen wrote:
> On Mon, Jun 18, 2012 at 6:35 PM, Simon Riggs wrote:
>> On 13 June 2012 19:28, Andres Freund wrote:
>>> This adds a new configuration parameter multimaster_node_id which determines
>>> the id used for wal originating in one cluster.
>>
>> Looks good and
On Tue, Jun 19, 2012 at 5:33 PM, Greg Smith wrote:
> In January of 2011 Robert committed 7f242d880b5b5d9642675517466d31373961cf98
> to try and compact the fsync queue when clients find it full. There's no
> visible behavior change, just a substantial performance boost possible in
> the rare but e
Robert Haas writes:
> On Tue, Jun 19, 2012 at 9:47 AM, Tom Lane wrote:
>> I thought the general idea was to use int32 most places, but int4 in
>> catalog declarations. I don't think it's tremendously important if
>> somebody uses the other though.
> I concur with Peter that TMTOWTDI is not the
On Tue, Jun 19, 2012 at 3:18 PM, Andres Freund wrote:
> More seriously: Even if we don't put MM in core I think putting the basis for
> it in core so that somebody can build such a solution reusing the existing
> infrastructure is a sensible idea. Imo the only thing that requires explicit
> suppor
On Tue, Jun 19, 2012 at 5:39 PM, Robert Haas wrote:
> On Tue, Jun 19, 2012 at 5:33 PM, Greg Smith wrote:
>> In January of 2011 Robert committed 7f242d880b5b5d9642675517466d31373961cf98
>> to try and compact the fsync queue when clients find it full. There's no
>> visible behavior change, just a
Excerpts from Robert Haas's message of mar jun 19 17:39:46 -0400 2012:
> On Tue, Jun 19, 2012 at 5:33 PM, Greg Smith wrote:
> > In January of 2011 Robert committed 7f242d880b5b5d9642675517466d31373961cf98
> > to try and compact the fsync queue when clients find it full. There's no
> > visible be
On Tue, Jun 19, 2012 at 11:59 AM, Robert Haas wrote:
> On Tue, Jun 19, 2012 at 1:56 AM, Tom Lane wrote:
> > The transaction would be committed before a command success report is
> > delivered to the client, so I don't think delivered-and-not-marked is
> > possible.
>
> ...unless you have configu
On 19 June 2012 14:03, Tom Lane wrote:
> "Every WAL record"? Why in heck would you attach it to every record?
> Surely putting it in WAL page headers would be sufficient. We could
> easily afford to burn a page switch (if not a whole segment switch)
> when changing masters.
This does appear to
Tom Lane wrote:
> I have not looked to see how many places do that. If it's a
reasonably
> small number of places, I'm OK with getting rid of int4 at the C
level.
> (int2/int8 the same of course.)
$ find -name '*.h' -or -name '*.c' | egrep -v '/tmp_check/' | xargs cat
\
| egrep -c '\bint2\
Excerpts from Josh Kupershmidt's message of mié may 30 14:55:12 -0400 2012:
> Hi all,
>
> Bosco Rama recently complained[1] about not seeing a message printed
> by pg_restore for each LO to be restored. The culprit seems to be the
> different level passed to ahlog() for this status message:
>
>
On Tue, Jun 19, 2012 at 5:46 PM, Robert Haas wrote:
>> Btw, what do you mean with "conflating" the stream? I don't really see that
>> being proposed.
>
> It seems to me that you are intent on using the WAL stream as the
> logical change stream. I think that's a bad design. Instead, you
> should
On 20 June 2012 05:46, Robert Haas wrote:
> It seems to me that you are intent on using the WAL stream as the
> logical change stream. I think that's a bad design. Instead, you
> should extract changes from WAL and then ship them around in a format
> that is specific to logical replication.
Th
On Tuesday, June 19, 2012 10:58:44 PM Marko Kreen wrote:
> On Mon, Jun 18, 2012 at 6:35 PM, Simon Riggs wrote:
> > On 13 June 2012 19:28, Andres Freund wrote:
> >> This adds a new configuration parameter multimaster_node_id which
> >> determines the id used for wal originating in one cluster.
> >
"Kevin Grittner" writes:
> Tom Lane wrote:
>> I have not looked to see how many places do that. If it's a reasonably
>> small number of places, I'm OK with getting rid of int4 at the C level.
>> (int2/int8 the same of course.)
> $ find -name '*.h' -or -name '*.c' | egrep -v '/tmp_check/' | xar
On 20 June 2012 04:31, Kevin Grittner wrote:
> I've done a lot of MM replication,
> and so far have not had to use a topology which allowed loops.
The proposal is to use WAL to generate the logical change stream. That
has been shown in testing to be around x4 faster than having a
separate change
1 - 100 of 130 matches
Mail list logo
