On 5 Aug 2011, at 1:05am, john darnell wrote:
> As a matter of fact, Simon, for some reason, whenever I call it in this
> particular function (and I call this function a lot), it returns an error.
> The error is (both surprisingly and unsurprisingly) "unable to close due to
> unfinalized
> Can you
> make absolutely sure sqlite3_close() has been called correctly and does not
> return
> an error ?
As a matter of fact, Simon, for some reason, whenever I call it in this
particular function (and I call this function a lot), it returns an error. The
error is (both surprisingly and
The table with the first five columns is the R*tree table, and thus already has
an index. I just tried adding an index to col5 also, but no change in
performance.
~Seth
On Aug 4, 2011, at 3:27 PM, Simon Slavin wrote:
>
> On 4 Aug 2011, at 9:59pm, Seth Price wrote:
>
>> Those fields span
On 4 Aug 2011, at 11:02pm, john darnell wrote:
> The documentation says that when I close a database transactions in progress
> are rolled back, but I cannot find a way of testing for whether a transaction
> is completed.
SQLite does not do any tasks in the background. There's no need to
>
> I'm not sure where you close your database handle, and what you mean by 'end
> of
> a run', but when you have used sqlite3_close() to close all handles to a
> SQLite
> database that file should no longer exist. If you still have a file with
> that name on
> your disk, something has gone
On 4 Aug 2011, at 9:59pm, Seth Price wrote:
> Those fields span two tables.
Then create one index for each table.
Simon.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Those fields span two tables.
~Seth
via iPhone
On Aug 4, 2011, at 2:45 PM, Simon Slavin wrote:
>
> On 4 Aug 2011, at 8:40pm, Seth Price wrote:
>
>> SELECT class FROM data_r JOIN data USING (rowNum) WHERE ?1 < col0min AND
>> col0max < ?2 AND ?3 < col1min AND col1max <
On 4 Aug 2011, at 8:40pm, Seth Price wrote:
> SELECT class FROM data_r JOIN data USING (rowNum) WHERE ?1 < col0min AND
> col0max < ?2 AND ?3 < col1min AND col1max < ?4 AND ?5 < col2min AND col2max <
> ?6 AND ?7 < col3min AND col3max < ?8 AND ?9 < col4min AND col4max < ?10 AND
> ?11 < col5 AND
On 4 Aug 2011, at 8:28pm, john darnell wrote:
> The name of the DB file I use is IndexData.db. On certain versions (but not
> all versions) of the plugin, during the processing, an IndexData.db-Journal
> file is created. It is always empty at the end of a run.
I'm not sure where you close
On Thu, Aug 4, 2011 at 9:44 PM, Alexey Pechnikov wrote:
>
> http://sqlite.mobigroup.ru/artifact/265e408b4352d66cfc79a9990cb2c22fb390d3b6
>
> http://sqlite.mobigroup.ru/artifact/2250bbbc83f80eff73ce003ab7a30293c688ae9b
That is one of the coolest things i've seen months
2011/8/4 Stephan Beal :
> i'm wondering if anyone can point me to an example of implementing such a
> beast?
http://sqlite.mobigroup.ru/artifact/265e408b4352d66cfc79a9990cb2c22fb390d3b6
http://sqlite.mobigroup.ru/artifact/2250bbbc83f80eff73ce003ab7a30293c688ae9b
P.S. Many
2011/8/4 Vinoth raj :
> So, the requirement is to save sqlite database on a server from a C++
> application.
There are a lot of ways to copy file to server. SQLite database is single file.
And you can send SQL dump. And you can send diff of SQL dump.
And you can export your
Hello people.
I apologize for the vagueness of this email, but all I can really hope for is
some ideas to pursue, I think.
I have an InDesign plug-in that scans numerous InDesign documents (there are no
limits, but a common number would be around 100) for certain names and stores
that
At 19:58 04/08/2011, Stephan Beal wrote:
>Hi, all!
>
>http://www.sqlite.org/whentouse.html
>
>says:
>
>"Temporary triggers can be added to the database to record all changes into
>a (temporary) undo/redo log table. These changes can then be played back
>when the user presses the Undo and Redo
On Thu, Aug 4, 2011 at 1:58 PM, Stephan Beal wrote:
> Hi, all!
>
> http://www.sqlite.org/whentouse.html
>
> says:
>
> "Temporary triggers can be added to the database to record all changes into
> a (temporary) undo/redo log table. These changes can then be played back
>
Removing "COUNT(*) AS count" and "GROUP BY class" and doing it in-program
shaved ~10% off of the time. I'll keep it. :)
~Seth
On Aug 4, 2011, at 11:30 AM, Eduardo Morras wrote:
>
> Oks, another let's try another thing/think.
>
> Try the select without the COUNT(*):
>
> SELECT class FROM
On Thu, Aug 4, 2011 at 12:52 PM, Stephan Beal wrote:
> On Thu, Aug 4, 2011 at 7:50 PM, Stephan Beal wrote:
>
>> http://www.sqlite.org/whentouse.html
>>
>
> Specifically: 2nd section, 1st list item.
It may depend on the usage. After reading about
On Thu, Aug 4, 2011 at 8:13 PM, Dustin Sallings wrote:
>https://github.com/dustin/kvtest/blob/master/sqlite-base.cc#L160
>
The code is short and to the point. Thank you :).
--
- stephan beal
http://wanderinghorse.net/home/stephan/
On 4 Aug 2011, at 18:15, Vinoth raj wrote:
> I have been using SQlite database since three years. Suddenly I have a
> requirement for client/server support for my project.
> So, the requirement is to save sqlite database on a server from a C++
> application.
We use sqlite in a client/server
On 4 Aug 2011, at 6:58pm, Stephan Beal wrote:
> http://www.sqlite.org/whentouse.html
>
> says:
>
> "Temporary triggers can be added to the database to record all changes into
> a (temporary) undo/redo log table. These changes can then be played back
> when the user presses the Undo and Redo
Hi, all!
http://www.sqlite.org/whentouse.html
says:
"Temporary triggers can be added to the database to record all changes into
a (temporary) undo/redo log table. These changes can then be played back
when the user presses the Undo and Redo buttons. Using this technique, an
unlimited depth
On Thu, Aug 4, 2011 at 7:50 PM, Stephan Beal wrote:
> http://www.sqlite.org/whentouse.html
>
Specifically: 2nd section, 1st list item.
--
- stephan beal
http://wanderinghorse.net/home/stephan/
___
sqlite-users mailing list
On Thu, Aug 4, 2011 at 7:15 PM, Vinoth raj wrote:
> It would be a great help if you can shed some light on my problem. Is it
> possible at all to save SQLite database on a server?
>
http://www.sqlite.org/whentouse.html
--
- stephan beal
Oks, another let's try another thing/think.
Try the select without the COUNT(*):
SELECT class FROM data_r JOIN data USING (rowNum) WHERE 57 < col0min
AND col0max < 61 AND 52 < col1min AND col1max < 56 AND 66 < col2min
AND col2max < 70 AND 88 < col3min AND col3max < 92 AND 133 < col4min
AND
Dear SQLite users,
I have been using SQlite database since three years. Suddenly I have a
requirement for client/server support for my project.
So, the requirement is to save sqlite database on a server from a C++
application.
I explored the SQLite API with no success. Even numeours queries on
I tried one index on all columns originally, but it didn't help nearly as much
as the R*tree.
~Seth
On Aug 4, 2011, at 10:26 AM, Simon Slavin wrote:
>
> On 4 Aug 2011, at 5:13pm, Seth Price wrote:
>
>> They were all simple indices on one or more columns, so if you have an idea
>> on a more
On 4 Aug 2011, at 5:13pm, Seth Price wrote:
> They were all simple indices on one or more columns, so if you have an idea
> on a more complex index, I'd apply it and test it out.
That's not going to be much use, is it ? A single SELECT which tests all those
different columns can only use one
> Any threads or processes would still be in contention for the same resource:
> access to that file on disk.
Midway through the run, `iostat` is showing me exactly 0 disk accesses, so I'm
assuming it's a locking problem and not a physical I/O bottleneck. I'm hoping
that you folks could give
> With SQLITE_THREADSAFE=2
> with SQLITE_THREADSAFE=1
With 2 threads and THREADSAFE=2 I get 167 seconds.
With 2 threads and THREADSAFE=1, I get 177 seconds.
With 1 thread and THREADSAFE=1, I get 53 seconds.
With 1 thread and THREADSAFE=2, I get 52 seconds.
One thing I'm worried about is I have
On 4 Aug 2011, at 4:44pm, Seth Price wrote:
> I was hoping that SQLite was doing that with the JOIN statement. It's
> narrowing down the results with the query from the R*table (col*min &
> col*max). A large query will have ~14k rows at this point. Then JOINs with
> the original table (data)
On 8/4/2011 11:35 AM, Sean Hammond wrote:
> Hey, I've been recording timestamped log messages in sqlite3 by using
> datetime('now') in INSERT queries, e.g.:
>
> INSERT INTO Logs (...,time) VALUES (...,datetime('now'));
>
> (The time column has type DATETIME.)
That's irrelevant. You are storing
On Aug 4, 2011, at 1:54 AM, Eduardo Morras wrote:
>
> They block each other not trying to get a lock, but trying to get
> access to disk and cache. Disk access time and cache is shared
> between all threads and if all threads needs access to different
> parts of the db they will figth like
> You didn't show your timing results or say what kind of machine you're
> running on.
I'm running on a 2.26 GHz Mac Pro with 8 physical cores and 16 GB of RAM.
> SQL error (635): near "ORDER": syntax error
You're probably seeing that error because you need to recompile with
Hey, I've been recording timestamped log messages in sqlite3 by using
datetime('now') in INSERT queries, e.g.:
INSERT INTO Logs (...,time) VALUES (...,datetime('now'));
(The time column has type DATETIME.)
I noticed that if I retrieve these rows with a SELECT query the
datetimes only have
Also..
.
Your elapsed time is using clock() which tells you processor time.
With SQLITE_THREADSAFE=2
59.855 - 1 thread
49.535 - 2 threads
92.789 - 3 threads
with SQLITE_THREADSAFE=1
61.146 - 1 thread
49.568 - 2 threads
64.932 - 3 threads
The way you're splitting your work is bad.
You didn't show your timing results or say what kind of machine you're running
on.
I'm also seeing
Calculating Subset Sample...
SQL error (635): near "ORDER": syntax error
I'm running Linux, sqlite 3.7.5, E5520 2.27Ghz Intel with 16 threads possible.
Here's my timings just on the first few rows
> my program language is Java. And the www.sqlite.org not supply API for
Java. What should I do?
I think this one is the best
http://code.google.com/p/sqlite-jdbc/
Best regards,
Sylvain
___
sqlite-users mailing list
sqlite-users@sqlite.org
At 06:53 04/08/2011, you wrote:
>I was hoping they wouldn't block each other because it's a read
>lock. I tried making an index on all the columns, but R*tree table +
>JOIN that I'm using runs about 10x faster. I might have done
>something wrong, so I'm open to suggestions on a better index,
38 matches
Mail list logo