Iker Arizmendi <[EMAIL PROTECTED]> wrote:
> Using pysqlite2 I noticed (using ps) that whenever I connected to
> a database for the first time an extra thread is spawned. At first
> I thought this might be the doing of pysqlite but a quick test with
> the sqlite3 command line tool confirmed that
My version of Linux is still using LinuxThreads (as opposed to NPTL) -
I guess that extra thread is the "manager" thread.
Iker
On 7/12/06, Iker Arizmendi <[EMAIL PROTECTED]> wrote:
Using pysqlite2 I noticed (using ps) that whenever I connected to
a database for the first time an extra thread
To write to the database use a single thread with a single DB instance
operating sync'd on a mutex to force exclusive write. Reads can be
anytime, provided they wait on the mutex and set it as a read lock.
You might find that to let writes queue until either an alarm triggers
or the queue
Using pysqlite2 I noticed (using ps) that whenever I connected to
a database for the first time an extra thread is spawned. At first
I thought this might be the doing of pysqlite but a quick test with
the sqlite3 command line tool confirmed that this occurs within
sqlite3 itself:
$ sqlite3
Jay gave you the clue. You need to have software and hardware which
will address data beyond 32 bits, i.e. 64 bit file addressing. The 64
bit file version of all the tools you mention should perform your job.
Now you choose according to the scale of your application, Enterprise,
Cesar David Rodas Maldonado wrote:
Hello to everybody!
I want to ask to the people something... If i have a table with about
8.000.000.000 registers, and i have to do a select in a numeric row that
will have index. what db do you recommend to me for use?
Thanxs (:D) people!
How many
At 19:59 12/07/2006, you wrote:
32 bits, the size of the key, only holds numbers up to 2 or 4 billion.
You said you have 8 billion rows.
You need a 64 bit version or to split up the data into many tables.
On 7/12/06, Cesar David Rodas Maldonado <[EMAIL PROTECTED]> wrote:
So Jay... Do you
I'm a bit stuck on this one problem
I have this simple table that has two fields: id and value.
What I want to do is to get the largest id (the last entry) and then add an
arbitrary number to the value that's in the same entry as that one, and put
the result in the next entry.
to make it sound
On 7/12/06, John Newby <[EMAIL PROTECTED]> wrote:
Hi, how can I find out the names of the fields within a given table?
I've tried "pragma table_info(test);"
but this brings back too much info, I just require the names as I'll be
storing them in an array within my application.
When you use
On Wed, 12 Jul 2006 19:05:51 +0100, you wrote:
> Hi, how can I find out the names of the
> fields within a given table?
>
> I've tried "pragma table_info(test);"
> but this brings back too much info,
> I just require the names as I'll be
> storing them in an array within my application.
Your
On 7/12/06, Gussimulator <[EMAIL PROTECTED]> wrote:
Yeah and while we're at it, he could invest a few bucks on a couple twin
diesel generators, a battery room, and whatnot, a group of hookers to save
the night.
You forgot the beer and pizza delivery contract!
Gussimulator No entendi tu mensaje, podrias porfavor escribirmelo en español
On 7/12/06, Gussimulator <[EMAIL PROTECTED]> wrote:
Yeah and while we're at it, he could invest a few bucks on a couple twin
diesel generators, a battery room, and whatnot, a group of hookers to save
the night.
Yeah and while we're at it, he could invest a few bucks on a couple twin
diesel generators, a battery room, and whatnot, a group of hookers to save
the night.
- Original Message -
From: "Jay Sprenkle" <[EMAIL PROTECTED]>
To:
Sent: Wednesday, July 12, 2006
Use a queue for the database operations in this case. You won't suffer from
lock or busy errors, if all access is serialized. Queues can scale very well
if done right.
Mike
-Ursprüngliche Nachricht-
Von: Gussimulator [mailto:[EMAIL PROTECTED]
Gesendet: Mittwoch, 12. Juli 2006 20:55
An:
Yes but I'd rather move to another database solution then.
As I see it, sqlite could well have such system built-in, no, it wouldnt be
bloat.
The queue system will work, as I mentioned I already thought of it, but, I
don't think its an elegant solution.
I could avoid the "can't read while
I cannot delay my thread.
Also, if I made a thread per event, then that wouldnt work either, since
every thread would open the database to try a write on it.
We are talking about 100 events in a second. This thread handles the queries
as it gets the events.
- Original Message -
On 7/12/06, Cesar David Rodas Maldonado <[EMAIL PROTECTED]> wrote:
I don't care about hardware... i will do cluster... but which DB motor do
think will be good for work with great velocity in select (sqlite,
postgresql or mysql).
ah.. i almost forgot... i will do a great number of select...
I
On 7/12/06, Pedro J. Vivancos <[EMAIL PROTECTED]> wrote:
I'm developing a program using C++ and SQLite3 as database. I have a
problem with encoding when I try to execute a SQL query with some
characters like "ñ, á, é, í, ó, ú". My db is UTF-8 encoded and when I'm
debugging I can see that the
I don't care about hardware... i will do cluster... but which DB motor do
think will be good for work with great velocity in select (sqlite,
postgresql or mysql).
ah.. i almost forgot... i will do a great number of select...
On 7/12/06, Jay Sprenkle <[EMAIL PROTECTED]> wrote:
32 bits, the size
Hi, how can I find out the names of the fields within a given table?
I've tried "pragma table_info(test);"
but this brings back too much info, I just require the names as I'll be
storing them in an array within my application.
Many thanks
John
32 bits, the size of the key, only holds numbers up to 2 or 4 billion.
You said you have 8 billion rows.
You need a 64 bit version or to split up the data into many tables.
On 7/12/06, Cesar David Rodas Maldonado <[EMAIL PROTECTED]> wrote:
So Jay... Do you think that i have to use a PC of 64
On 7/12/06, Gussimulator <[EMAIL PROTECTED]> wrote:
About the "cant read while writing", how to avoid this?, I cant stop my
system while using the database. There would be no point on having a
database then.
check the return code from operation, if it says Busy, or Locked,
the redo the
I would suggest using a queue approach. Its not as bad as it sounds. Have
that queue processed by a single thread, which controls the database writes.
That way you can save the time for other operations, such as database
open/close etc.
You can't avoid the "can't read while writing" though.
So Jay... Do you think that i have to use a PC of 64 bits, or 32bits pc with
clustering?
On 7/12/06, Jay Sprenkle <[EMAIL PROTECTED]> wrote:
> Hello to everybody!
>
> I want to ask to the people something... If i have a table with about
> 8.000.000.000 registers, and i have to do a select in
Dear friends,
I'm developing a program using C++ and SQLite3 as database. I have a
problem with encoding when I try to execute a SQL query with some
characters like "ñ, á, é, í, ó, ú". My db is UTF-8 encoded and when I'm
debugging I can see that the sentence is right but I have no result.
Yes, in theory.
However, when about 100 queries are fired in a second, only a few get the
data populated..
I can not tell when to start or end the transactions, because events are
fired from a third-party system. So I used a timer, set to 5 seconds.. then
tried with 10 seconds, got the same
Hello to everybody!
I want to ask to the people something... If i have a table with about
8.000.000.000 registers, and i have to do a select in a numeric row that
will have index. what db do you recommend to me for use?
Thanxs (:D) people!
Iker Arizmendi <[EMAIL PROTECTED]> wrote:
> When a statement of the form:
>
> CREATE TABLE xxx IF NOT EXISTS
>
> what sort of locking overhead is incurred in the case
> where the table already exists?
>
If the table already exists, no locking occurs. The
statement above is a no-op.
--
D.
When a statement of the form:
CREATE TABLE xxx IF NOT EXISTS
what sort of locking overhead is incurred in the case
where the table already exists?
Regards,
Iker
--
Iker Arizmendi
e: [EMAIL PROTECTED]
w: ikerariz.blogspot.com
Hello to everybody!
I want to ask to the people something... If i have a table with about
8.000.000.000 registers, and i have to do a select in a numeric row that
will have index. what db do you recommend to me for use?
Thanxs (:D) people!
On 7/12/06, Gussimulator <[EMAIL PROTECTED]> wrote:
Hello,
I'm currently using SQLite3 on my multi-threaded software.
I have tried several ways for dealing with my issue, however, I came to the
conclusion that there must be some trick I havent been told of.
It wasn't clear to me when I
First of, thank you very much Christian for replying so precisely to
my message.
Things to check:
- Multiple processes accessing the database file. SQLite will arbitrate
access using a well defined locking protocol. If a process does not
respect this protocol, or locking is defective (NFS
Hello,
I'm currently using SQLite3 on my multi-threaded software.
I have tried several ways for dealing with my issue, however, I came to the
conclusion that there must be some trick I havent been told of.
Heres the situation, Theres a big amount of input coming in from the software
to the
On Wed, 2006-07-12 at 01:12 +0100, Nuno Lucas wrote:
> On 7/12/06, Gregary Hendricks <[EMAIL PROTECTED]> wrote:
> > Greetings fellow SQLiters,
> >
> > I am having a bad hair day as a result of an error I am getting with my
> > SQLite database.
> >
> > I am running SQLite 3.3.6 on SuSE linux 9.3
>
thanks! I was able to compile sqlite for the arm architecture. I followed
the directions here:
http://www.sqlite.org/cvstrac/wiki?p=HowToCompile=1070624069
but modified it so it looks like this:
unzip sqlite_source.zip
rm tclsqlite.c
for i in *.c; do arm-linux-gcc -O -c $i; done
rm shell.o
Christian Smith wrote:
Keiichi McGuire uttered:
This is a bit confusing, especially since if I compile it using gcc it
will
compile w/o any problems. What would this "incompatibility" mean and
what
would a solution be to make it compatible and be found by the compiler?
Chances are that
Getting an executable requires two steps. First you compile the
application program then you link the output of the compiler with
standard and special link libraries.
I am not familiar with your setup and don't know what Linux distribution
you are using or machine you are running but seeing
The error message you got says that libsqlite.a is incompatible. GCC is
working fine, it is the link libraries which are the problem. If you
get a source release of Sqlite, run "configure" and compile it then
install you will have a known environment.
Keiichi McGuire wrote:
This is a bit
Sorry, Hopefully someone else can give you advice to get past this. I
will say the time I had an "incompatible" library, I was using a binary
copy that had been cross-compiled. My solution was to take that library
(which was otk) and compile it directly on my linux box. Why that made a
difference
> Never mind. I just realized (from rereading a previous newsgroup
> posting) that the wildcards can only be used to replace values within
> expressions (and not the table or column names).
>
>
I use printf for these cases before preparing:
sprintf(buf, "INSERT INTO %s VALUES (?,
Hi,
Ralf Junker wrote:
> 2. xBestIndex / xFilter / sqlite3_index_info
>
> The sqlite3_index_info structure contains char *idxStr; which may be used to
> the application's liking. In addition, int needToFreeIdxStr; can be used to
> instruct SQLite3 to free idxStr.
>
> Would it be possible
Am 12.07.2006 um 11:16 schrieb RohitPatel:
INT and INTEGER behaves differently (for PRIMARY KEY) !!!
(SQLite 3.3.4)
create table t1 ( id INTEGER PRIMARY KEY );
create table t2 ( id INT PRIMARY KEY );
insert into t1 values(NULL);
insert into t1 values(NULL);
insert into t2 values(NULL);
Bull219 uttered:
Dear all,
I am developping a freeware which uses SQLite. One of my beta testers
informed me about an issue he had: with his DB, following the query which is
sent to the DB, I have the error in the subject of this email. I did some
testing, and when I succeeded in
Keiichi McGuire uttered:
This is a bit confusing, especially since if I compile it using gcc it will
compile w/o any problems. What would this "incompatibility" mean and what
would a solution be to make it compatible and be found by the compiler?
Chances are that the libsqlite.a you're
INT and INTEGER behaves differently (for PRIMARY KEY) !!!
(SQLite 3.3.4)
create table t1 ( id INTEGER PRIMARY KEY );
create table t2 ( id INT PRIMARY KEY );
insert into t1 values(NULL);
insert into t1 values(NULL);
insert into t2 values(NULL);
insert into t2 values(NULL);
/* insert into t1
Thank you very much for clearing my doubt.
Rohit.
--
View this message in context:
http://www.nabble.com/id-INTEGER-PRIMARY-KEY--vs--PRIMARY-KEY-%28id%29-tf1922943.html#a5285302
Sent from the SQLite forum at Nabble.com.
46 matches
Mail list logo