It doesn't collect those statistics automatically, as part of query
plan optimization?
- Original Message -
From: "Dan Kennedy"
To: "General Discussion of SQLite Database"
Sent: Friday, August 14, 2009 11:37 PM
Subject: Re: [sqlite] Multiple indexes in SQLite, and selecting which
to u
Thanks for the feedback. I realise that Dan's suggestions won't necessarily
fix the problem, but it would be very handy to know if my theory about poor
choice of indexes is right. So, Dan, take your bow!
- Original Message -
From: "Dan Kennedy"
To: "General Discussion of SQLite Database
On Aug 15, 2009, at 1:34 PM, Jim Showalter wrote:
> How will that help him fix this problem, if the problem is that
> SQLite's query optimizer is selecting a suboptimal index to use, and
> there is no way to specify which index to use?
The statistics collected by the ANALYZE command will be used
How will that help him fix this problem, if the problem is that
SQLite's query optimizer is selecting a suboptimal index to use, and
there is no way to specify which index to use?
- Original Message -
From: "Dan Kennedy"
To: "General Discussion of SQLite Database"
Sent: Friday, August
On Aug 15, 2009, at 1:08 PM, His Nerdship wrote:
>
> Good day,
> We have a puzzling problem with a large (1GB+) database.
> Most of our queries are based on 3 columns, say X, Y and Z.
> X is always the first in the index. However, sometimes the query
> involves a
> small range of Y and a large
Good day,
We have a puzzling problem with a large (1GB+) database.
Most of our queries are based on 3 columns, say X, Y and Z.
X is always the first in the index. However, sometimes the query involves a
small range of Y and a larger range of Z, and sometimes the reverse. We
first had an index ba
On Aug 15, 2009, at 2:14 AM, Ken wrote:
> I'm not sure if this an issue or not. make test failed with the
> following:
>
> 2 errors out of 40872 tests
> Failures on these tests: rollback-2.3 tkt3457-1.4
> All memory allocations freed - no leaks
> Memory used: now 0 max 1
On Aug 15, 2009, at 9:32 AM, cscs-sql...@usa.net wrote:
>
> Did an FTS3 update change how many negation operators (dash/-) can
> be used in
> a
> match statement?
>
> For example, in sqlite3.dll version 3.5.7:
>
> colname match 'tetons -bend -jackson -oxbow* -parks' works as
> expected; bend,
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Sebastian Arcus wrote:
> Hi and thanks for the suggestion. I did as you advised and ran 'vmstat
> 1' in a terminal. Very little activity - maybe 20-40kb every 6-7 seconds
> on the bo - pretty much nothing on bi. Also, zeros all around on so/si.
Tha
Did an FTS3 update change how many negation operators (dash/-) can be used in
a
match statement?
For example, in sqlite3.dll version 3.5.7:
colname match 'tetons -bend -jackson -oxbow* -parks' works as expected; bend,
jackson, oxbow* and parks are all removed from the results.
but
With, sqlite
Ricardo Ayres Severo
wrote:
> My question is if I would face any problem by opening the database
> when my application starts and actually never closing it. I don't do
> any access across threads.
You shouldn't have any problems with this. It's a pretty common access
pattern (except that most ap
Roger Binns wrote:
> Sebastian Arcus wrote:
>
>> The SQLite documentation talks about entire database locks by operations
>> of the order of milliseconds - 10 seconds seems a long way off.
>>
>
> There is a possible but unlikely cause for what you are seeing. In order to
> ensure the dat
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Sebastian Arcus wrote:
> The SQLite documentation talks about entire database locks by operations
> of the order of milliseconds - 10 seconds seems a long way off.
There is a possible but unlikely cause for what you are seeing. In order to
ensure t
Hi all,
I'm using the mozStorage implementation from Mozilla foundation, with
Javascript and XUL (for the UI) to build a business-type app.
I'm having this problem where I seem to get a lock of about 10 seconds
after a read operation on a table, before I can perform a write
operation on the sa
On Fri, 14 Aug 2009 11:00:53 -0700, Dmitri Priimak
wrote:
>Kees Nuyt wrote:
>> On Fri, 14 Aug 2009 07:24:31 -0700, Dmitri Priimak
>> wrote:
>>
>>> Hi.
>>>
>>> I have a database with few simple tables. Database is updated regularly
>>> and than distributed to the clients, which only use it for
On Fri, 14 Aug 2009 12:33:30 -0400, Angus March wrote:
> I want my INSERT done right away, I just don't want it to be flushed
> from the filesystem's write-behind cache until the kernel decides, not
> when SQLite decides.
Did you mean you do "want it to be flushed from the filesystem's
write-
I'm not sure if this an issue or not. make test failed with the following:
2 errors out of 40872 tests
Failures on these tests: rollback-2.3 tkt3457-1.4
All memory allocations freed - no leaks
Memory used: now 0 max 102680 max-size2800336
Page-cache used: now
Hi All,
I'm using SQLite3 with a Embedded Linux application. The problem I'm
facing is that there's a bug in the Linux kernel I'm using that the
open files counter isn't being decremented when a file descriptor is
closed. With that my application simply crashes after some time of
usage.
Because of
Kees Nuyt wrote:
> On Fri, 14 Aug 2009 07:24:31 -0700, Dmitri Priimak
> wrote:
>
>> Hi.
>>
>> I have a database with few simple tables. Database is updated regularly
>> and than distributed to the clients, which only use it for reading. So,
>> concurrency is not an issue. But the database is al
On Fri, 14 Aug 2009 07:24:31 -0700, Dmitri Priimak
wrote:
>Hi.
>
>I have a database with few simple tables. Database is updated regularly
>and than distributed to the clients, which only use it for reading. So,
>concurrency is not an issue. But the database is already quite large.
>The file is
Thanks, a lot guys.
I will run a few tests with dummy database populated with my target amount
rows and will let you know results. I realize now that I seem to have some
unfounded fear of large files (FOLF) :) Hopefully it will pass.
--
Dmitri Priimak
___
On 14 Aug 2009, at 5:33pm, Angus March wrote:
> I want my INSERT done right away,
Then do not turn off synchronous !
> I just don't want it to be flushed
> from the filesystem's write-behind cache until the kernel decides, not
> when SQLite decides.
SQLite cannot control how your operating sys
Simon Slavin wrote:
> On 14 Aug 2009, at 5:25pm, Angus March wrote:
>
>
>> I need to know that if I turn of the synchronous that no synching will
>> be done, up to, and including, when the session is closed. I'm asking,
>> because my program just INSERTs once per session, so if a synch gets
>> d
On 14 Aug 2009, at 5:25pm, Angus March wrote:
> I need to know that if I turn of the synchronous that no synching will
> be done, up to, and including, when the session is closed. I'm asking,
> because my program just INSERTs once per session, so if a synch gets
> done when the session closes, th
I need to know that if I turn of the synchronous that no synching will
be done, up to, and including, when the session is closed. I'm asking,
because my program just INSERTs once per session, so if a synch gets
done when the session closes, that's pretty useless.
___
> Your only problem is that you're at Stanford and Dr Hipp was at
> Duke so he hates you.
Not much humor on this list, but you made my day :-)
Very funny.
-Clark
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mai
On 14 Aug 2009, at 3:24pm, Dmitri Priimak wrote:
> I have a database with few simple tables. Database is updated
> regularly
> and than distributed to the clients, which only use it for reading.
> So,
> concurrency is not an issue. But the database is already quite large.
> The file is about
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Michael Han wrote:
> My question is can I safely turn off this flag if the underlying file
> system supports journaling?
By default journaling file systems only use their journal for meta-data
(eg directory entries and allocation information). They
On Fri, Aug 14, 2009 at 07:24:31AM -0700, Dmitri Priimak scratched on the wall:
> I have a database with few simple tables. Database is updated regularly
> and than distributed to the clients, which only use it for reading. So,
> concurrency is not an issue. But the database is already quite lar
On Fri, Aug 14, 2009 at 9:24 AM, Dmitri Priimak wrote:
> Hi.
>
> I have a database with few simple tables. Database is updated regularly
> and than distributed to the clients, which only use it for reading. So,
> concurrency is not an issue. But the database is already quite large.
> The file is ab
Hi.
I have a database with few simple tables. Database is updated regularly
and than distributed to the clients, which only use it for reading. So,
concurrency is not an issue. But the database is already quite large.
The file is about 2.6GB and one table has about 1,800,000 rows
and another on
On Fri, Aug 14, 2009 at 05:10:40PM +0800, Michael Han scratched on the wall:
> My question is can I safely turn off this flag if the underlying file
> system supports journaling?
No.
As the docs state, with synchronous=ON the system verifies the data
actually hits disk before proceeding.
A.J.Millan wrote:
> May be of interest to add in the doc, that, when the second argument
> is negative, "the the first character of the substring is found by
> counting from the right rather than the left", but remember that the
> last actual character is -2, because -1 is just the ending NULL.
se
On Fri, 14 Aug 2009, Dan Kennedy wrote:
>
> On Aug 14, 2009, at 5:14 PM, Hugh Sasse wrote:
>
> > The upshot of my tests on Solaris9 was:
> >
> > WARNING: Multi-threaded tests skipped: Linked against a non-
> > threadsafe Tcl build
> > All memory allocations freed - no leaks
> > Memory used:
On Aug 14, 2009, at 5:14 PM, Hugh Sasse wrote:
> The upshot of my tests on Solaris9 was:
>
> WARNING: Multi-threaded tests skipped: Linked against a non-
> threadsafe Tcl build
> All memory allocations freed - no leaks
> Memory used: now 0 max5727984 max-size
> 1717987
The upshot of my tests on Solaris9 was:
WARNING: Multi-threaded tests skipped: Linked against a non-threadsafe Tcl build
All memory allocations freed - no leaks
Memory used: now 0 max5727984 max-size 171798754
Page-cache used: now 0 max 13 max-size
Hi List:
May be of interest to add in the doc, that, when the second argument is
negative, "the the first character of the substring is found by counting
from the right rather than the left", but remember that the last actual
character is -2, because -1 is just the ending NULL.
Of course it is
Hello,
My application is using SQLite and I find a huge discrepancy of
committing speed when "synchronous" flag is set on / off (via PRAGMA).
My question is can I safely turn off this flag if the underlying file
system supports journaling? The FAQ
(http://www.sqlite.org/faq.html#q5) said "Another
38 matches
Mail list logo