Digital Workshop

Is it possible for Opus to use the DSN to locate an sqlite database file when the publication is published.

I would like to be able to read the content of an sqlite data file on a server from a publication that is installed on a client machine. Is this possible?

Thanks

Opus Pro 6.0.1

Bump. Anyone?

( Seems like there might be six or so questions in there. )

Perhaps I should clarify my initial posting.

I have a publication which is currently run on a customers server. The Opus publication connects to an sqlite database. The database is located in the <SYSTEM_PUBLICATION_FOLDER> on the server. The client machines on the network run the opus application from the server using shortcuts.

The problem is that the opus publication is running very slowly and there is a lag time of about 1 minute which causes a 100% CPU usage whenever anything is attempted in the publication. The server is a Quad core Intel Xeon 1.6 with 4GB of RAM. The Network is 100mbs+

The publication works perfectly when run from the server. The problem is only experienced when run accross the network.

I have come to the conclusion after many hours of testing that the only solution is to have the publication installed on each client machine on the network and have the data read from and written to the sqlite database on the server.

Is there any way to get Opus to connect to the database on the server from a client computer using ODBC. It seems that opus ignores the System DSN when the publication is published and will only read from the "Path for file when published" setting.

How can I get the publication installed on a client machine to read from the database on the server using ODBC.

Clearly this is a matter of urgency, as my customer needs a solution yesterday. Any help or advice would be appreciated.

Ken,

So I take it the server is simply a basic file server. (if it were configured as a web server, you could use GetInternetData/PostInternetData.)

Have you done a Search on the forum? Check generally for 'database' and later specific 'sqlite'.

My thinking is that if your SQLite database is in a specific and static location on a local LAN (such as: j:databases\usershared\xyz) then maybe it is just a matter of plugging that path into whatever MEANS of connection to the DB you choose (DSN, ODBC...). Issue may be whether the shared drives are mounted automatically.

If you insist on using ODBC are you using the "SQLite specific ODBC"? My understanding is that the connection settings must be configured on each client machine (via control panel on Vista).

Hence... the benefit of connecting without ODBC. Look at some of the other Posts here. There is definitely a way to connect to a SQLite DB without a client machine ODBC. (still I think the key is whether the path to the file is 'static' / 'fixed'.)

Anyone else got any ideas here?

SQLite was designed as a single user database.

Browse to the official SQLite.org site, see page http://sqlite.org/lockingv3.html
Scroll down to Section 6.0 "How To Corrupt Your Database Files".
You will see "Your best defense is to not use SQLite for files on a network filesystem."
Personally I would use a DBMS like MySQL for this kind of situation, or possibly access the database using a server side language on the same machine as the database (maybe PHP using PDO if DB is SQLite v3).

Paul