We use connection pooling. It took only a few lines of C code to implement.
In our case we maintain a global list of open connections and access
them by pathname. A mutex is required to protect the list anchor
pointer when a new connection is added.
We also add pthread read/write locking to the shared connection which
lets us cut the record locking out of Sqlite, of special benefit in an
embedded application.
Dinesh Ramdass wrote:
HI ,
*How we use SQlite:*
We use MONO and SQLite 3 for our project which runs on Linux. It is
going to be used as an embedded database in a device.
*What i want to do:*
I want to use connection pooling either using some feature provided by
SQLite or someone else
*What i want to know:*
I read in SQLite FAQs that we can have multiple connections. But there
was a limitation with using same connection across threads. I think that u
were referring to use of Connection pooling. So my question is around this
*Question 1:* Does SQLite direct provide support for connection pooling in
form of any library or opensource binary?
*Question 2:* If there is no direct support for connection pooling, then is
it becos we should not use? Does SQLite not advise the use of implementing
connection pooling?
*Question 3:* If my database access is going to be often but only for 5-6
rows CRUD operation at a time, then do u advice me to open a new connection
and close it after use instead of maintaing an open connection for a long
time or resetting a connection using timeout?
Kindly reply A.S.A.P. Its pretty urgent .
*Also i need to give a justification for not implementing CONNECTION
POOLING
to my client.*
Kinldy provide me a justification.
be Happy,
Dinesh.
-----------------------------------------------------------------------------
To unsubscribe, send email to [EMAIL PROTECTED]
-----------------------------------------------------------------------------
