Daniel Wood writes:
> Is the main difference between:
> if (s->blockState != TBLOCK_SUBINPROGESS)
> elog(*FATAL*, ...
> vs
> Assert(s->blockState == TBLOCK_SUBINPROGRESS);
> the fact that in both cases:
> a) the situation is unexpected, as in no user code can create this;
These two questions are about the correct coding practice in Postgresql
vs the specifics of xact.c
Is the main difference between:
if (s->blockState != TBLOCK_SUBINPROGESS)
elog(*FATAL*, ...
vs
Assert(s->blockState == TBLOCK_SUBINPROGRESS);
the fact that in both cases:
a) t
It seems the backend and libpq don't agree. The backend makes no
special provision to wait for a CopyDone message if an error occurs
during copy-both. It simply sends an ErrorResponse and that's it.
libpq, on the other hand, treats either CopyDone or ErrorResponse as a
cue to transition to PGASYN
On Fri, Apr 26, 2013 at 05:52:03PM -0700, Jeff Davis wrote:
> On Fri, 2013-04-26 at 16:46 -0700, David Fetter wrote:
> > On Fri, Apr 26, 2013 at 07:49:47PM +, Tom Lane wrote:
> > > Given this risk and the lack of field complaints about the issue, it
> > > doesn't seem prudent to back-patch.
>
On Fri, 2013-04-26 at 16:46 -0700, David Fetter wrote:
> On Fri, Apr 26, 2013 at 07:49:47PM +, Tom Lane wrote:
> > Given this risk and the lack of field complaints about the issue, it
> > doesn't seem prudent to back-patch.
...
> This needs back-patching to 9.1, where the bug was introduced.
On Sat, 2013-04-27 at 00:20 +0200, Andres Freund wrote:
> CFLAGS_VECTORIZATION? EXTRA sounds to generic to me.
I went with CFLAGS_VECTOR to be a little shorter while still keeping
some meaning.
> I think it would be better to have a PGAC_PROG_CC_VAR_OPT or so which
> assigns the flag to some pass
On Fri, Apr 26, 2013 at 07:49:47PM +, Tom Lane wrote:
> Fix collation assignment for aggregates with ORDER BY.
>
> ORDER BY expressions were being treated the same as regular aggregate
> arguments for purposes of collation determination, but really they should
> not affect the aggregate's coll
On 2013-04-26 12:57:09 -0700, Jeff Davis wrote:
> I updated the patch and split it into two parts (attached).
> The second patch adds the configure-time check for the right
> compilation flags, and uses them when compiling checksum.c. I
> called the new variable CFLAGS_EXTRA, for lack of a better
After 87306184580c9c49717, if the postmaster dies without cleaning up (i.e.
power outage), running "pg_ctl start" just gives this message and then
exits:
pg_ctl: another server might be running
Under the old behavior, it would try to start the server anyway, and
succeed, then go through recovery
On Fri, 2013-04-26 at 16:40 -0400, Greg Smith wrote:
> I think I need to do two baselines: master without checksums, and
> master with extra optimizations but still without checksums. It may be
> the case that using better compile time optimizations gives a general
> speedup that's worth consi
On 4/26/13 3:57 PM, Jeff Davis wrote:
The second patch adds the configure-time check for the right
compilation flags, and uses them when compiling checksum.c. I
called the new variable CFLAGS_EXTRA, for lack of a better idea,
so feel free to come up with a new name. It doesn't check for, or
use,
On 27/04/13 02:48, Tom Lane wrote:
Robert Haas writes:
On Thu, Apr 25, 2013 at 1:56 PM, Tom Lane wrote:
However, the documentation in libpq.sgml is a bit bogus too, because it
counsels trying the PQputCopyEnd() call again, which will not work
(since we already changed the asyncStatus). We co
r
use, -msse4.1, but that can be specified by the user by
configuring with CFLAGS_EXTRA="-msse4.1".
I don't know of any more required changes, aside from
documentation improvements.
Regards,
Jeff Davis
fnv-jeff-20130426.patch
Description: Binary data
fnv-jeff-20130426-cfl
David Fetter writes:
> While testing the upcoming FILTER clause for aggregates, Erik Rijkers
> uncovered a long-standing bug in $subject, namely that this case
> wasn't handled. Please find attached a patch by Andrew Gierth and
> myself which fixes this issue and adds a regression test to ensure
Joe Conway writes:
> Committed back to 9.1
Thanks,
--
Dimitri Fontaine
http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-ha
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 04/08/2013 08:34 AM, Dimitri Fontaine wrote:
> Joe Conway writes:
>> OK, maybe I'll try to take a look in the meantime.
>
> That would be awesome :)
>
>> Did you have any comment on the other pg_dump patch (reviewed by
>> Vibhor)?
>
> This whole
On 26.04.2013 19:50, Magnus Hagander wrote:
On Fri, Apr 26, 2013 at 6:43 PM, Simon Riggs wrote:
On 26 April 2013 17:25, Heikki Linnakangas wrote:
Actually, from a usability point of view I think would be nice to have just
one setting, "recovery_target". It's already somewhat confusing to have
On Fri, Apr 26, 2013 at 6:43 PM, Simon Riggs wrote:
> On 26 April 2013 17:25, Heikki Linnakangas wrote:
>> On 26.04.2013 19:05, Simon Riggs wrote:
>>>
>>> On 26 April 2013 16:38, Robert Haas wrote:
On Fri, Apr 26, 2013 at 11:35 AM, Simon Riggs
wrote:
>
> Given that I was d
On 26 April 2013 17:25, Heikki Linnakangas wrote:
> On 26.04.2013 19:05, Simon Riggs wrote:
>>
>> On 26 April 2013 16:38, Robert Haas wrote:
>>>
>>> On Fri, Apr 26, 2013 at 11:35 AM, Simon Riggs
>>> wrote:
Given that I was describing how we might implement Heikki's
suggestion, I fi
On Fri, Apr 26, 2013 at 12:25 PM, Heikki Linnakangas
wrote:
>> Doing it the other way means you need to add a new kind of recovery
>> target to the API just for this.
>> recovery_target_immediate = on
>
> Sounds good to me.
Yeah, I don't have a problem with that, at all.
> Actually, from a usabi
Ashutosh Bapat writes:
> On Fri, Apr 26, 2013 at 7:54 PM, Tom Lane wrote:
>> A larger point is that the patch as proposed doesn't fix the stated
>> problem, because it only descends into written-out subqueries. It
>> would only succeed at looking into views if we applied it after
>> rewriting, r
On Fri, Apr 26, 2013 at 2:08 AM, Andres Freund wrote:
> On 2013-04-25 23:07:02 -0400, Peter Eisentraut wrote:
> > I'm not sure who is supposed to be able to read this sort of stuff:
> >
> > Latest checkpoint's NextXID: 0/7575
> > Latest checkpoint's NextOID: 49152
> > Latest chec
On 26.04.2013 19:05, Simon Riggs wrote:
On 26 April 2013 16:38, Robert Haas wrote:
On Fri, Apr 26, 2013 at 11:35 AM, Simon Riggs wrote:
Given that I was describing how we might implement Heikki's
suggestion, I find this comment confusing.
Please explain.
Heikki's suggestion is simply to ha
On 26 April 2013 16:38, Robert Haas wrote:
> On Fri, Apr 26, 2013 at 11:35 AM, Simon Riggs wrote:
>> Given that I was describing how we might implement Heikki's
>> suggestion, I find this comment confusing.
>>
>> Please explain.
>
> Heikki's suggestion is simply to have a mode that stops as soon
On Thu, Apr 25, 2013 at 12:17 PM, Tom Lane wrote:
> The cause is that each one will wait for all older snapshots to be
> gone --- and it does that before dropping its own snapshot, so that the
> other ones will see it as something to be waited out too.
This makes sense. Thank you for explaining.
On Thu, Apr 25, 2013 at 12:09 AM, Tom Lane wrote:
> Alvaro Herrera writes:
>> Jeff Janes escribió:
>>> With the stats file split patch 187492b6c2e8cafc5 introduced in 9.3dev, now
>>> after a crash the postmaster will try to delete all files in the directory
>>> stats_temp_directory. When that is
On Fri, Apr 26, 2013 at 11:35 AM, Simon Riggs wrote:
> Given that I was describing how we might implement Heikki's
> suggestion, I find this comment confusing.
>
> Please explain.
Heikki's suggestion is simply to have a mode that stops as soon as
consistency is reached. The server already knows
On 26 April 2013 15:38, Robert Haas wrote:
> On Fri, Apr 26, 2013 at 10:05 AM, Simon Riggs wrote:
>> Restore points are definitely the way to go here, this is what they
>> were created for. Stopping at a labelled location has a defined
>> meaning for the user, which is much better than just "stop
On Fri, Apr 26, 2013 at 7:54 PM, Tom Lane wrote:
> Ashutosh Bapat writes:
> > The reason being, it doesn't look into the subqueries (in FROM clause) to
> > infer that p.product_id is essentially product.product_id which is a
> > primary key.
>
> Right.
>
> > Attached find a crude patch to infer
On Fri, Apr 26, 2013 at 10:48 AM, Tom Lane wrote:
> Robert Haas writes:
>> On Thu, Apr 25, 2013 at 1:56 PM, Tom Lane wrote:
>>> However, the documentation in libpq.sgml is a bit bogus too, because it
>>> counsels trying the PQputCopyEnd() call again, which will not work
>>> (since we already cha
I explain more detail about this problem.
This problem was occurred by RestartPoint create illegal WAL file in during
archive recovery. But I cannot recognize why illegal WAL file was created
in CreateRestartPoint(). My attached patch is really plain…
In problem case at XLogFileReadAnyTLI(), fir
On Apr 26, 2013 4:38 PM, "Robert Haas" wrote:
>
> On Fri, Apr 26, 2013 at 10:05 AM, Simon Riggs
wrote:
> > Restore points are definitely the way to go here, this is what they
> > were created for. Stopping at a labelled location has a defined
> > meaning for the user, which is much better than ju
Robert Haas writes:
> On Thu, Apr 25, 2013 at 1:56 PM, Tom Lane wrote:
>> However, the documentation in libpq.sgml is a bit bogus too, because it
>> counsels trying the PQputCopyEnd() call again, which will not work
>> (since we already changed the asyncStatus). We could make that say "a
>> zero
On Fri, Apr 26, 2013 at 10:05 AM, Simon Riggs wrote:
> Restore points are definitely the way to go here, this is what they
> were created for. Stopping at a labelled location has a defined
> meaning for the user, which is much better than just "stop anywhere
> convenient", which I found so frighte
On Thu, Apr 25, 2013 at 1:56 PM, Tom Lane wrote:
> However, the documentation in libpq.sgml is a bit bogus too, because it
> counsels trying the PQputCopyEnd() call again, which will not work
> (since we already changed the asyncStatus). We could make that say "a
> zero result is informational, y
Ashutosh Bapat writes:
> The reason being, it doesn't look into the subqueries (in FROM clause) to
> infer that p.product_id is essentially product.product_id which is a
> primary key.
Right.
> Attached find a crude patch to infer the same by traversing subqueries.
I think this is probably a ba
On 26 April 2013 14:40, Tom Lane wrote:
> Andres Freund writes:
>>> On 2013-04-26 13:11:00 +0200, Florian Pflug wrote:
The unresolved code issue that I know of is moving the compiler flags
behind a configure check. I would greatly appreciate it if you could take
a look at that. M
On 26 April 2013 14:48, Tom Lane wrote:
> Magnus Hagander writes:
>> That said, maybe the easier choice for a *system* (such as v-thingy)
>> would be to simply to the full backup using pg_basebackup -x (or
>> similar), therefor not needing the log archive at all when restoring.
>> Yes, it makes t
Hi All,
If group by clause has primary key, the targetlist may have columns which
are not part of the aggregate and not part of group by clause. The relevant
commit is e49ae8d3bc588294d07ce1a1272b31718cfca5ef and relevant mail thread
has subject Functional dependencies and GROUP BY.
As a result, f
Magnus Hagander writes:
> That said, maybe the easier choice for a *system* (such as v-thingy)
> would be to simply to the full backup using pg_basebackup -x (or
> similar), therefor not needing the log archive at all when restoring.
> Yes, it makes the base backup slightly larger, but also much
>
Andres Freund writes:
>> On 2013-04-26 13:11:00 +0200, Florian Pflug wrote:
>>> The unresolved code issue that I know of is moving the compiler flags
>>> behind a configure check. I would greatly appreciate it if you could take a
>>> look at that. My config-fu is weak and it would take me some t
On 26.04.2013 14:54, Magnus Hagander wrote:
On Fri, Apr 26, 2013 at 1:47 PM, Simon Riggs wrote:
On 26 April 2013 11:29, Heikki Linnakangas wrote:
But there is also an operation to simply "restore a backup".
Just because a tool supports an imprecise definition of a restore,
doesn't mean Pos
On Fri, Apr 26, 2013 at 5:08 AM, Andres Freund wrote:
> I have to admit I don't see the point. None of those values is particularly
> interesting to anybody without implementation level knowledge and those
> will likely deal with them just fine. And I find the version with the
> shorter names far
On 26 April 2013 12:54, Magnus Hagander wrote:
> That said, maybe the easier choice for a *system* (such as v-thingy)
> would be to simply to the full backup using pg_basebackup -x (or
> similar), therefor not needing the log archive at all when restoring.
> Yes, it makes the base backup slightly
On Fri, Apr 26, 2013 at 1:47 PM, Simon Riggs wrote:
> On 26 April 2013 11:29, Heikki Linnakangas wrote:
>
>> But there is also an operation to simply "restore a backup".
>
> Just because a tool supports an imprecise definition of a restore,
> doesn't mean Postgres should encourage and support tha
On 26 April 2013 11:29, Heikki Linnakangas wrote:
> But there is also an operation to simply "restore a backup".
Just because a tool supports an imprecise definition of a restore,
doesn't mean Postgres should encourage and support that.
"Restore a backup" is more suited to filesystems where mos
--On 25. April 2013 23:19:14 -0400 Tom Lane wrote:
I think I've heard of scripts grepping the output of pg_controldata for
this that or the other. Any rewording of the labels would break that.
While I'm not opposed to improving the labels, I would vote against your
second, abbreviated scheme
On 2013-04-26 13:11:00 +0200, Florian Pflug wrote:
> > The unresolved code issue that I know of is moving the compiler flags
> > behind a configure check. I would greatly appreciate it if you could take a
> > look at that. My config-fu is weak and it would take me some time to figure
> > out how
On Apr26, 2013, at 10:28 , Ants Aasma wrote:
> On Apr 25, 2013 10:38 PM, "Jeff Davis" wrote:
> >
> > On Tue, 2013-04-23 at 11:44 +0300, Ants Aasma wrote:
> > > I will try to reword.
> >
> > Did you have a chance to clarify this, as well as some of the other
> > documentation issues Simon mentione
On 26.04.2013 12:16, Simon Riggs wrote:
On 18 April 2013 19:11, Heikki Linnakangas wrote:
I just found out that if you use continuous archiving and online backups,
it's surprisingly difficult to restore a backup, without replaying any more
WAL than necessary.
I didn't add it myself because I
On 18 April 2013 19:11, Heikki Linnakangas wrote:
> I just found out that if you use continuous archiving and online backups,
> it's surprisingly difficult to restore a backup, without replaying any more
> WAL than necessary.
I didn't add it myself because I don't see the need, if we think more
On 2013-04-25 23:07:02 -0400, Peter Eisentraut wrote:
> I'm not sure who is supposed to be able to read this sort of stuff:
>
> Latest checkpoint's NextXID: 0/7575
> Latest checkpoint's NextOID: 49152
> Latest checkpoint's NextMultiXactId: 7
> Latest checkpoint's NextMultiOffset
On 26/04/13 18:53, Daniel Farina wrote:
On Thu, Apr 25, 2013 at 9:34 PM, Tom Lane wrote:
Alvaro Herrera writes:
Tom Lane wrote:
I think I've heard of scripts grepping the output of pg_controldata for
this that or the other. Any rewording of the labels would break that.
While I'm not opposed
Hi,
I discavered the problem cause. I think taht horiguchi's discovery is another
problem...
Problem has CreateRestartPoint. In recovery mode, PG should not WAL record.
Because PG does not know latest WAL file location.
But in this problem case, PG(standby) write WAL file at RestartPoint in arch
On Apr 25, 2013 10:38 PM, "Jeff Davis" wrote:
>
> On Tue, 2013-04-23 at 11:44 +0300, Ants Aasma wrote:
> > I will try to reword.
>
> Did you have a chance to clarify this, as well as some of the other
> documentation issues Simon mentioned here?
>
> http://www.postgresql.org/message-id/CA+U5nMKVEu
On 26.04.2013 07:47, Amit Langote wrote:
How would code after applying this patch behave if a recycled segment gets
renamed using the newest timeline (say 3) while we are still recovering from
a lower timeline (say 2)? In that case, since XLogFileReadAnyTLI returns
that recycled segment as the
What would happen if a recycled segment gets assigned a newer timeline than
the one we are currently recovering from? In the reported erroneous
behavior, that is what happens causing XLogFileReadAnyTLI() to return such
bogus segment causing the error. Since, expectedTLIs contains a newer
timeline a
Peter Eisentraut writes:
> It seems pretty straightforward and useful, so I'm not sure where your
> hesitation is coming from.
If you're talking about my hesitation to consider what we have now as
marketing material worthy, it comes from the fact that I don't have a
use case where I don't need sp
On 26.04.2013 07:02, Kyotaro HORIGUCHI wrote:
I am uncertain a bit weather it is necessary to move curFileTLI to
anywhere randomly read . On a short glance, the random access occurs also
for reading checkpoint-related records.
I didn't understand that.
Also I don't have clear distinction bet
59 matches
Mail list logo