Sébastien Lardière wrote:
On 26/08/2009 04:46, Mark Kirkwood wrote:
Sébastien Lardière wrote:
Hi All,
I've a cluster ( Pg 8.3.7 ) with WAL Shipping, and a few hours ago,
the master had to restart.
I use walmgr from Skytools, which works very well.
I have already restart the master wi
Sébastien Lardière wrote:
Hi All,
I've a cluster ( Pg 8.3.7 ) with WAL Shipping, and a few hours ago,
the master had to restart.
I use walmgr from Skytools, which works very well.
I have already restart the master without any problem, but today, the
slave doesn't work like I want. The field
Should have mentioned : assuming you are on a platform where you *have*
a choice about compilation word-length!
(Solaris and ?)
Mark Kirkwood wrote:
Now suppose you want to run a Pg database for such a situation may
as well compile 32-bit.
---(end of
No disagreement from me about the 64-bit *hardware* and *os*...
Now suppose you want to run a Pg database for such a situation may
as well compile 32-bit.
Why ? well you *dont* want to set shared_buffers to 20G... in fact 200M
works better -
why ? well your 64-bit os file cache is much more
Wouldn't you only care about 64-bit Postgres if you wanted to make
shared_buffers bigger than 4G?
Various other posters have commented about the sweet-spot for
shared_buffers being ~ 100-200M (or thereabouts).
So it seems to me that there is nothing to be gained using a 64-bit
binary with the
FYI - Ext3 has 3 modes :
data=ordered(default) : metadata is journaled (at write time data is
written before metadata - i.e ordered)
data=journal: data and metadata are journaled
data=writeback: metadata journaled (no ordering at write time)
The default will not help to protect database integrit
Along similar lines - have generally obtained better server performance
(and stability) from most Linux distros after replacing their supplied
kernel with one from kernel.org .
regards
Mark
Josh Berkus wrote:
Folks,
While debugging a wireless card, I came across this interesting bit:
http://
Tom Lane wrote:
The fact that RI triggers issue SQL commands is an artifact of
their implementation (and one that I believe Stephan and Jan would like
to get rid of); they shouldn't be cluttering the log at all.
I am glad you mentioned that - I did find myself wondering why it was
necessary t
This caught me today :
I switched on "log_statement=true" whilst examining a possible foreign
key concurrency problem. I noticed that the generated foreign key check
"SELECT 1 FROM ONLY ... WHERE id = ...FOR UPDATE..."
on the parent table seemed to be only appearing every now and again.
This c
I would suspect some *other* service is using the 4G for transient
storage every now and again, and it just so happens that Pg is getting
tripped up.
What else does this machine run ?
regards
Mark
Nigel J. Andrews wrote:
On Fri, 9 Jan 2004, Aurangzeb M. Agha wrote:
Here's the output of
I just tried :
i) "wal" in "7.4 docs" -> 5 seconds
ii) "partial index" in "All Sites" -> 3 seconds
iii) "column statistics" in "All Sites" -> 2 seconds
iv) "fork exec" in "All Sites" -> 2 seconds
These results seem pretty good to me...
regards
Mark
---(end of broad
Tom Lane wrote:
Not sure that it's fair to characterize this as a property of the
relational model. It is a property of the SQL standard.
Yes indeed - I fell into the classic "Relational model and SQL are not
the same thing" trap !
Mark
---(end of broadcast)
Might be worth trying a larger statistics target (say 100), in the hope
that the planner then has better information to work with.
best wishes
Mark
Marc G. Fournier wrote:
he problem is that right now, we look at the LIKE first, giving us ~300k
rows, and then search through those for those who
ftp://ftp.au.postgresql.org/pub/postgresql/ is still only displaying
7.4.0 right now...
Marc G. Fournier wrote:
Send the details to webmaster@ so that the guys can add it to the database
... but, how are you mirroring? All official mirrors use rsync, and they
are auto-added to the list of mirro
You should see a lines like this in configure.in
[lines 28-33 of configure.in for 7.2.4]
AC_PREREQ(2.13)
AC_CONFIG_AUX_DIR(config)
VERSION='7.2.4'
AC_SUBST(VERSION)
AC_DEFINE_UNQUOTED(PG_VERSION, "$VERSION")
best wishes
Mark
bpalmer wrote:
I'm trying to figure out what version of a source code
Jeff wrote:
UNless the controller itself has a battery backed cache it is dangerous - there
are many more failures than losing power. Ie, blowing out the power supply or cpu.
We've burnt up a fair share of cpu's over the years. Luckly on a Sun it isn't that
big a deal.. but on x86. wel... yo
Furthermore, if the disk drives are lying to the controller, it's
anybody's guess whether or not data ever actually gets to the disk.
When is it safe to let blocks expire out of the controller cache?
If your computer can't know if the data has been written (because of
drives that lie), I can't im
Bjørn T Johansen wrote:
Just a question... Are there any reasons not to just take the source and
compile it under RHEL 3.0? Or am I missing something?
(We are about to install 3.0, so I would really like to know..)
In fact there are some reasons *to* do this :
- compiler optimizations specific
Maybe it is a little late to be posting on this thread - but I was doing
pgbench runs with a Raid 0 ATA system and thought the results might be
interesting.
So here they are : pgbench -c 5 -t 1000 -s 5, median of 3 runs on a
Dual PIII 700 512Mb 2x7200 RPM ATA 133 Promise TX200
(same method / Pg
Dave Weaver wrote:
- clustering the "obs" table on "station"
Sorry, I don't understand what you mean by this - can you explain?
Supposing obs_pkey is on (station, valid_time):
cluster obs_pkey on obs
will re-order the rows in obs based on the index obs_pkey. (This is
clustering on 'sta
scott.marlowe wrote:
Was there a performance difference in the set with write cache on or off?
Yes - just in the process of a little study concerning this - I will
post some preliminary results soon
cheers
Mark
---(end of broadcast)---
TIP 5
Got to going this today, after a small delay due to the arrival of new
disks,
So the system is 2x700Mhz PIII, 512 Mb, Promise TX2000, 2x40G ATA-133
Maxtor Diamond+8 .
The relevent software is Freebsd 4.8 and Postgresql 7.4 Beta 2.
Two runs of 'pgbench -c 50 -t 100 -s 10 bench' with a power
Its worth checking - isn't it ?
I appeciate that you may have performed such tests previously - but as
hardware and software evolve its often worth repeating such tests (goes
away to do the suggested one tonight).
Note that I am not trying to argue away the issue about write caching -
it *has*
Mark Kirkwood wrote:
I should have said that I was using Freebsd 4.8 with write caching off.
write caching *on* - I got myself confused about what the value "1"
means
---(end of broadcast)---
TIP 9: the planner will ignore your
Some sort of ATA Raid is probably worth considering -
e.g. I am experimenting with a system using 2 ATA-66 Seagates + 1
Promise TX2000
The disks themselves give fairly poor performance when attached to the
std IDE channels :
sequential write 15Mb/s
sequential read 20Mb/s
But attached to the Pr
Most web appl servers have settings to do this for you, or APIs you can
use.
e.g. for J2EE see SessionBindingListener
regards
mark
Egor Shipovalov wrote:
Is there a way to make transaction automatically COMMIT or ROLLBACK after
certain period of time?
I want website visitors to be able to pag
Could be the the database is checkpointing then. Try experimenting with :
checkpoint_segments
checkpoint_timeout
Might be worth playing with :
wal_buffers
as well
regards
Mark
u15074 wrote:
The performance is ok and stays constant over the whole time. But I have the
following effect: from ti
I built and ran Pg on Tru64 about a year ago.
I recall having to edit pg_ctl a bit...
Does that count as experience?
regards
Mark
Richard Huxton wrote:
On Thursday 19 Jun 2003 12:23 pm, Justin Clift wrote:
Hi everyone,
Does someone feel like saying Hi to Brad here, and assisting him with
Is using Solaris 8 an option ?
Postgresql 7.1.1, 7.1.2 and 7.2dev all seem to run ok on this release of
Solaris.
( I used the 64 bit enabled kernel and upped shmmax)
regards
Mark
---(end of broadcast)---
TIP 1: subscribe and unsubscribe command
> > I tried this query :
> >
> > SELECT sum(val) FROM fact0
> >
> > for Postgres, Db2 and Oracle. The results were
> >
> > Postgres2m25s
> > Db2 40s
> > Oracle 50s
> >
> > This seems to be the likely culprit. I suspect that the "many
> > block/page read at once" type optimzations
Previously:
>
>When attempting to build JDBC 7.0 driver with "make jdbc2 jar", the
following
>error occurs:
>
> org.postgresql.Connection.java: Can't find class
org/postgresql/Field
>
>Any ideas? Thanks.
I encountered this error when using an older JDK ( < 1.2 )
Downloading 1.2.2 from S
Ross,
I am not so sure that ALTER family is used for development
purposes only...
Consider data warehousing where typically each day the administrator
would :
1Extract data from source database(s) - usually in ascii file
format, as these dbs may be connection unfriendly, not controlled by
32 matches
Mail list logo
