John Smith wrote:
Architecture: Intel Core 2 Duo
OS: linux-2.6.20-gentoo-r8
Filesystem: ext3
Postgres v8.2.3 compiled with gcc 4.1.1-r3
RAM - 2GB
Shared buffers - 24MB
[All other Postgres configuration parameters are default values]
Problem description:
COPY into temp table fails using a
I remember Heikki mentioned improving redo recovery in one of the
emails in the past, so I know people are already thinking about this.
I have some ideas and just wanted to get comments here.
ISTM that its important to keep the redo recovery time as small as possible
in order to reduce the
On Thu, 2008-02-28 at 14:38 +0530, Pavan Deolasee wrote:
I had this idea sometime back. Not sure if this has been discussed before
Check the archives for my post to hackers in Jan 2007 and subsequent
discussion. It's possible, just a little fiddly.
--
Simon Riggs
2ndQuadrant
Heikki Linnakangas wrote:
Attached is a simple patch to fix that by disallowing
CREATE+DROP+PREPARE TRANSACTION more reliably.
That patch was missing changes to header files. New patch attached.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
Index:
Plausible theory, and nice explanation
Try the following link (I had to wait for 50 sec for the link to appear, but
I guess the trade-off of getting knowledge in return is worth it :) )
http://www5.upload2.net/download/77fa86e16a02e52fd5439c76e148d231/47c7fdce/rfsLfnuVlYjEcCJ/basetables.tgz
I'm now looking into toast code and I found following code in
pglz_decompress:
00704 if (destsize != source-rawsize)
00705 elog(destsize source-rawsize ? FATAL : ERROR,
00706 compressed data is corrupt);
I'm surprise why we there panic? By my opinion is not too good
* Pavan Deolasee:
The current redo-recovery is a single threaded, synchronous process.
The XLOG is read sequentially, each log record is examined and
replayed if required. This requires reading disk blocks in the
shared buffers and applying changes to the buffer. The reading
happens
Florian G. Pflug [EMAIL PROTECTED] writes:
I know that Simon has some ideas about parallel restored, though I don't
know how he wants to solve the dependency issues involved. Perhaps by
not parallelizing withon one table or index...
I think we should be *extremely* cautious about introducing
Alvaro Herrera [EMAIL PROTECTED] writes:
Zdenek Kotala wrote:
I'm now looking into toast code and I found following code in
pglz_decompress:
00704 if (destsize != source-rawsize)
00705 elog(destsize source-rawsize ? FATAL : ERROR,
00706 compressed data is
Pavan Deolasee wrote:
What I am thinking is if we can read ahead these blocks in the shared
buffers and then apply redo changes to them, it can potentially
improve things a lot. If there are multiple read requests, kernel (or
controller ?) can probably schedule the reads more efficiently.
Alvaro Herrera napsal(a):
Zdenek Kotala wrote:
I'm now looking into toast code and I found following code in
pglz_decompress:
00704 if (destsize != source-rawsize)
00705 elog(destsize source-rawsize ? FATAL : ERROR,
00706 compressed data is corrupt);
I'm surprise
Zdenek Kotala wrote:
I'm now looking into toast code and I found following code in
pglz_decompress:
00704 if (destsize != source-rawsize)
00705 elog(destsize source-rawsize ? FATAL : ERROR,
00706 compressed data is corrupt);
I'm surprise why we there panic?
Sergey Burladyan [EMAIL PROTECTED] writes:
[ xmlvalidate is a security hole ]
Given that this function is not documented nor tested in the regression
tests, I propose diking it out entirely.
regards, tom lane
---(end of
Tom Lane napsal(a):
Alvaro Herrera [EMAIL PROTECTED] writes:
Zdenek Kotala wrote:
I'm now looking into toast code and I found following code in
pglz_decompress:
00704 if (destsize != source-rawsize)
00705 elog(destsize source-rawsize ? FATAL : ERROR,
00706
Sometimes, I have to add column to table, then adding the column to some view.
Currently it impossible to alter a view to add a new column, which is really
annoying. The only solution is to drop the view, and all dependented view,
then recreating everything.
Someone on #postgresqlfr said me
Need help and direction creating new aggregate functions.
We need to add more average functions for both scientific and finical
purposes
RMS for electrical measurement purposes
Mode for both electrical and finical
Weighted Average finical purposes
Generalized mean for electrical
On Fri, Feb 29, 2008 at 12:11:59PM -0500, Justin wrote:
Need help and direction creating new aggregate functions.
We need to add more average functions for both scientific and finical
purposes
[ ... ]
what would be the best way to create these new functions??
I'd be tempted to look at
On Fri, 2008-02-29 at 15:49 +0100, Florian G. Pflug wrote:
I know that Simon has some ideas about parallel restored, though I don't
know how he wants to solve the dependency issues involved. Perhaps by
not parallelizing withon one table or index...
Well, I think that problem is secondary to
Jorgen Austvik - Sun Norway [EMAIL PROTECTED] writes:
Tom Lane wrote:
This is unfortunate and surprising, since that patch was intended to
prevent compilers from making unsafe alignment assumptions, but it sure
looks like this compiler has instead added a new one. Could you poke
into it ---
Peter Eisentraut escribió:
Am Freitag, 29. Februar 2008 schrieb Tom Lane:
Sergey Burladyan [EMAIL PROTECTED] writes:
[ xmlvalidate is a security hole ]
Given that this function is not documented nor tested in the regression
tests, I propose diking it out entirely.
Yes, it was
On Feb 29, 2008, at 8:10 AM, Florian Weimer wrote:
In the end, I wouldn't be surprised if for most loads, cache warming
effects dominated recovery times, at least when the machine is not
starved on RAM.
Uh... that's exactly what all the synchronous reads are doing...
warming the cache. And
Alvaro Herrera [EMAIL PROTECTED] writes:
Peter Eisentraut escribió:
Yes, it was accidentally left over from previous work. We should have
removed
it before the release, but that would have required an initdb.
So let's change it for a function that elog(ERROR)s on entry.
Yeah, I was just
This probably belongs on General, but
On Fri, Feb 29, 2008 at 9:11 AM, Justin [EMAIL PROTECTED] wrote:
Need help and direction creating new aggregate functions.
We need to add more average functions for both scientific and finical
purposes
RMS for electrical measurement purposes
Mode
Am Freitag, 29. Februar 2008 schrieb Tom Lane:
Sergey Burladyan [EMAIL PROTECTED] writes:
[ xmlvalidate is a security hole ]
Given that this function is not documented nor tested in the regression
tests, I propose diking it out entirely.
Yes, it was accidentally left over from previous
Decibel! wrote:
On Feb 29, 2008, at 8:10 AM, Florian Weimer wrote:
In the end, I wouldn't be surprised if for most loads, cache warming
effects dominated recovery times, at least when the machine is not
starved on RAM.
Uh... that's exactly what all the synchronous reads are doing... warming
Decibel! [EMAIL PROTECTED] writes:
Perhaps a good short-term measure would be to have recovery allocate
a 16M buffer and read in entire xlog files at once.
If that isn't entirely useless, you need a better kernel. The system
should *certainly* be bright enough to do read-ahead for our reads
* Tom Lane [EMAIL PROTECTED] [080229 15:49]:
If that isn't entirely useless, you need a better kernel. The system
should *certainly* be bright enough to do read-ahead for our reads of
the source xlog file. The fetches that are likely to be problematic are
the ones for pages in the data
Gurjeet Singh wrote:
Plausible theory, and nice explanation
Try the following link (I had to wait for 50 sec for the link to appear, but
I guess the trade-off of getting knowledge in return is worth it :) )
On Feb 26, 2008, at 4:36 PM, Tom Lane wrote:
I think a sane way to think about what Simon would like to accomplish
is not turn psql into a parallel job scheduler
My $0.02: I often find myself wishing I could perform parallel
operations in psql. There was a proposal for that that came up
Florian G. Pflug wrote:
The same holds true for index scans, though. Maybe we can find a
solution that benefits both cases - something along the line of a
bgreader process
I posted a patch to do readahead for bitmap index scans using posix_fadvise.
Experiments showed it works great on raid
Greg Stark wrote:
Florian G. Pflug wrote:
The same holds true for index scans, though. Maybe we can find a
solution that benefits both cases - something along the line of a
bgreader process
I posted a patch to do readahead for bitmap index scans using
posix_fadvise. Experiments showed it
31 matches
Mail list logo