"ITAGAKI Takahiro" <[EMAIL PROTECTED]> wrote
>
> AbsorbFsyncRequests will be called during the fsync loop in my patch,
> so new files might be added to pendingOpsTable and they will be removed
> from the table *before* writing the pages belonging to them.
> So I changed it to copy the contents of
"Qingqing Zhou" <[EMAIL PROTECTED]> wrote:
>while ((entry = (PendingOperationEntry *) hash_seq_search(&hstat)) != NULL)
>{
> + if (i >= count)
> +elog(ERROR, "pendingOpsTable corrupted");
> +
> + memcpy(&entries[i++], entry, sizeof(PendingOperationEntry));
> +
> + if (hash_search
"ITAGAKI Takahiro" <[EMAIL PROTECTED]> wrote
>
> Attached is a patch that fixes overflow of bgwriter's file-fsync request
> queue.
>
while ((entry = (PendingOperationEntry *) hash_seq_search(&hstat)) !=
NULL)
{
+ if (i >= count)
+elog(ERROR, "pendingOpsTable corrupted");
+
+ memcp
Attached is a patch that fixes overflow of bgwriter's file-fsync request queue.
It happened on heavy update workloads and the performance decreased.
I have sent HACKERS the detail.
---
ITAGAKI Takahiro
NTT Cyber Space Laboratories
bgwriter-requests-queue-overflow-2.patch
Description: Binary dat
Christopher Kings-Lynne <[EMAIL PROTECTED]> writes:
> With all this system view love going on, is there any point having a
> 'pg_savepoints' view to see what savepoints you've made?
Probably not, seeing that one of the main situations where you'd want to
know that would be after an error, and SEL
Attached is a patch that removes undesired paddings from b-tree indexes.
The tuples of b-tree index consist of BTItemData and index keys. BTItemData
should be placed only 2 byte alignment, so if the alignment of keys are less
than MAXIMUM_ALIGNOF, we can place them with their minimum alignment ins
I think it is worth distinguishing more clearly between portals that
should be displayed to the user and those that should not (which might
be labelled "internal cursors", perhaps). The tests above seem fairly
ad-hoc.
With all this system view love going on, is there any point having a
'pg_save
Neil Conway <[EMAIL PROTECTED]> writes:
> Why list Portals that cannot be directly manipulated by the user?
Define "directly manipulated". AFAIR there isn't any particular
distinction between portals that got made by DECLARE CURSOR and those
that got made by Bind; you can use either EXECUTE or Fe
On Thu, 2006-01-12 at 19:27 -0500, Tom Lane wrote:
> No kidding. But what do you think the correct test is? The comment's
> claim that stuff named "" should be suppressed seems
> wrong to begin with, as those are perfectly good cursors.
Well, I suggested to Joachim offlist that these shouldn't b
Neil Conway <[EMAIL PROTECTED]> writes:
> The tests above seem fairly ad-hoc.
No kidding. But what do you think the correct test is? The comment's
claim that stuff named "" should be suppressed seems
wrong to begin with, as those are perfectly good cursors. I'm also
unconvinced that a test on p
On Thu, 2006-01-12 at 10:51 +0100, Joachim Wieland wrote:
> Attached is a patch to create a new system view pg_cursors to list all
> available cursors to the current session and transaction.
+ /* loop until we find a named portal or hit the end of the list */
+ while ((hentry = hash_seq_search(has
On Wed, 2006-01-11 at 14:46 -0500, Neil Conway wrote:
> Barring any objections I'll apply this (only to HEAD) later today or
> tomorrow.
Applied.
-Neil
---(end of broadcast)---
TIP 3: Have you checked our extensive FAQ?
http://www
On Thu, 2006-01-12 at 00:28 -0500, Neil Conway wrote:
> Ah, right. Attached is a corrected patch.
Applied to HEAD, REL8_1_STABLE, REL8_0_STABLE, and REL7_4_STABLE.
-Neil
---(end of broadcast)---
TIP 4: Have you searched our list archives?
Folks,
Please find enclosed a doc patch that adds an example of a correllated
UPDATE.
Cheers,
D
--
David Fetter [EMAIL PROTECTED] http://fetter.org/
phone: +1 415 235 3778
Remember to vote!
Index: doc/src/sgml/ref/update.sgml
===
R
Mark Kirkwood wrote:
> Yes it would - I usually build the SGML -> HTML, then cut the code out
> of a browser session to test - the pain is waiting for the docs to build.
FWIW, what I do is to build a cut-down version of postgres.sgml to
include only the file you want to check. I think there is
Neil Conway <[EMAIL PROTECTED]> writes:
> Ah, I see what you mean. In implementing this, I wasn't sure the best
> way to provide these two sorts of TupleDesc references. My first thought
> was to add a "use ResourceOwner?" boolean parameter to the routines that
> create and destroy references to Tu
On Tue, 2006-01-10 at 18:05 -0500, Tom Lane wrote:
> Yeah. I was envisioning two different approaches: for TupleDesc
> references from long-lived data structures (ie, typcache or relcache)
> just increment or decrement the count when the referencing data
> structure changes. ResourceOwner would b
Attached is a patch to create a new system view pg_cursors to list all
available cursors to the current session and transaction.
The patch itself is quite analogous to the patch for pg_prepared_statements
I submitted in December.
The attributes is_holdable, is_binary and is_scrollable are exposed
18 matches
Mail list logo
