Dear Michael Fuhr,
The word you said is correct on most information. I will give you more information.
I'd guess that you haven't installed some third-party modules that you need on the Linux box, or that you've installed them in the wrong place. I don't make sure about third-party informat
On Mon, Feb 07, 2005 at 11:40:41AM +0700, Premsun Choltanwanich wrote:
>
> I found that at least Large Object that now failed and no data can be
> imported to Linux version. The problem I found was shown like:
>
> ERROR: could not access file "$libdir/lo": No such file or directory
> ERROR: co
On Sun, Feb 06, 2005 at 10:53:38PM -0500, Greg Stark wrote:
> This is a well understood problem. I remember it from my Systems class in
> school. And searching on google finds lecture notes that match my memory that
> there are other systems generally preferred to LRU precisely because they
> don't
For first information, I already have Large Object that work fine on PostgreSQL Windows Version. Now I plan to migrate all of Windows version to Linux version. But I got some problem that make it cannot work fine like on Windows version.
I found that at least Large Object that now failed and n
Greg Stark <[EMAIL PROTECTED]> writes:
> I recall the clock sweep having
> been recommended in class as having most of the best properties of LRU with
> very low cost in the critical path.
Right. The "pending move to front" idea that I suggested is basically a
variant of a clock algorithm: it tak
Tom Lane <[EMAIL PROTECTED]> writes:
> ReadBuffer needs to do a lookup to map the page ID to a buffer ID,
> which in principle requires only a shared lock on the page-to-buffer
> mapping (embodied in the buf_table hash table). Assuming success, it
> also needs to mark the buffer pinned and updat
Neil Conway <[EMAIL PROTECTED]> writes:
> We're only concerned with a buffer's access recency when it is on the
> free list, right?
Right; pinned buffers are not logically part of the freelist. (Whether
they are so physically is open to choice.)
Another free variable, AFAICS, is whether to do th
On Sun, 2005-02-06 at 19:30 -0500, Tom Lane wrote:
> This would be pretty good from a locking point of view, except that
> "update the LRU state" seems to require taking an exclusive lock on a
> global data structure, which puts us about back where we were.
We're only concerned with a buffer's acc
[EMAIL PROTECTED] writes:
> One of the things that is disturbing to me about the analyze settings is
> that it wants to sample the same number of records from a table regardless
> of the size of that table.
The papers that I looked at say that this rule has a good solid
statistical foundation, at
We've been speculating for awhile about ways to reduce contention for
the BufMgrLock, in hopes of fixing the "context swap storm" problem
exhibited by certain patterns of concurrent queries. The traditional
way to fix such problems is to divide what is protected by the lock
into finer-grain lockab
On Sun, 6 Feb 2005, Euler Taveira de Oliveira wrote:
Hi Tom,
Trolling -hackers for attachments would require a deal of manual
attention, unfortunately, because a lot of messages of that kind
would be bug reports not patches. I'm not real sure that there
aren't some non-patch attachments on -patche
Can anyone think of a reason we aren't inlining MemoryContextSwitchTo()
in GCC builds, similarly to the way list_head() et al are handled?
It wouldn't be a huge gain, but I consistently see MemoryContextSwitchTo
eating a percent or three of most profiles.
regards, tom lane
Hi Tom,
> Trolling -hackers for attachments would require a deal of manual
> attention, unfortunately, because a lot of messages of that kind
> would be bug reports not patches. I'm not real sure that there
> aren't some non-patch attachments on -patches as well.
>
That's why we need a Bug Trac
Marc G. Fournier wrote:
please let me know the URL for the message, so that I can see what it
was overlooked, and see if I can't improve the 'search' ...
Marc - here is one I submitted that was not picked up :-)
http://archives.postgresql.org/pgsql-patches/2005-01/msg00145.php
regards
Mark
---
On Sun, 6 Feb 2005, Greg Sabino Mullane wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Josh Berkus originally wrote:
Hey, for my information (people ask me this a lot) can one of you do a count
of patch submitters for 8.0? For 7.4, it was around 180.
Perhaps this is the gist of the problem:
This thread somehow reminds me how the company I'm working for
tries to measure my profitability: lines of code edited,
number of bugs fixed in what time, and other odd things :-)
Sorry if this is the wrong subject ...
Greg Sabino Mullane schrieb:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Josh Berkus originally wrote:
> Hey, for my information (people ask me this a lot) can one of you do a count
> of patch submitters for 8.0? For 7.4, it was around 180.
Perhaps this is the gist of the problem: the 180 count was not "patch
submitt
Ron Mayer <[EMAIL PROTECTED]> writes:
> What I think I'd like to see is for there to be
> another statistic similar to "correlation" but rather
> than looking at the total-ordering of the table, to
> look how correlated values within any single page are.
Our current approach to correlation is sure
Larry Rosenman writes:
> I had done the following, which I think is what's doing it:
> 1) alter table virtusers (and all the others in that db) set without oids;
Ah. I was just about to complain that I couldn't reproduce it, but
with that it crashes:
regression=# CREATE TABLE virtusers ...
regr
On Sun, 6 Feb 2005, Larry Rosenman wrote:
1) alter table virtusers (and all the others in that db) set without oids;
2) changed postgresql.conf's default_with_oids to false.
Based on my read, this case is what's causing the grief.
To get me out of it:
pg_dump exim >exim.db
psql template1
alter data
Euler Taveira de Oliveira <[EMAIL PROTECTED]> writes:
> You just consider the plain text attachments. What about the encoded
> attachments [1] and the gziped [2] ones?
And what of people who sent patches in-line (not as MIME-ified
attachments)? Or who sent them to -hackers instead of -patches?
T
Stephen Frost <[EMAIL PROTECTED]> writes:
> I guess I'm confused by 'who' needs to 'get in' to have a bit of code
> run at the very end of the backend startup. Apparently I'm somewhat
> naive in that area.
The point is that the postmaster is going to use the flat files to
check whether you're all
Short summary:
I had the same problem - since the sort order of zip-codes,
counties, city names, and states don't match, the optimizer
grossly overestimated the number of pages that would be read.
I bet doing a CLUSTER by ZIP would solve that particular
query, but would break similar queries b
On Sun, 6 Feb 2005, Larry Rosenman wrote:
I get the following:
$ TRAP: FailedAssertion("!(!(tup->t_data->t_infomask & 0x0010))", File:
"heapam.c", Line: 1133)
when I try to cluster this table:
CREATE TABLE virtusers (
lhs text,
rhs text,
insert_date timestamp(0) with time zone DEFAULT no
I get the following:
$ TRAP: FailedAssertion("!(!(tup->t_data->t_infomask & 0x0010))", File:
"heapam.c", Line: 1133)
when I try to cluster this table:
CREATE TABLE virtusers (
lhs text,
rhs text,
insert_date timestamp(0) with time zone DEFAULT now(),
insert_who text DEFAULT "curren
Neil Conway wrote:
Robert Treat wrote:
Actually i believe people want both syntax's as the former is used by
oracle and the latter by db2 (iirc)
I think the past consensus has been to adopt the SQL standard syntax. Is
there any reason to also support the Oracle syntax other than for
compatibili
26 matches
Mail list logo
