Besides my own experience, although it is a bit older and I couldn't find it in this email list, It is still in my archive, so there you have (copy-paste) of related email of 2014-09-08 in this list, send by jose isaias cabrera <...@cinops.xerox.com> in response of a query:
Re: [sqlite] Does the Connection string support UNC paths? a...@zator.com wrote... > >> >> ---- Mensaje original ---- >> De: "Chris" >> Para: >> Fecha: Sat, 6 Sep 2014 23:46:19 -0500 >> Asunto: [sqlite] Does the Connection string support UNC paths? >> >> >>I am old database programmer that just came across SQLite and am working >>on >>a small project for a PVR that uses SQLite as it's db provider. I try >>specifying a UNC path to the database for the datasource in the connection >>string and I get the following error, "unable to open database file". . >>When I look at the exception generated, I see an errorcode = 14. However, >>if I map a network drive, I can open the file and work with it. I am >>running Windows 7 x64 Pro and system.data.sqlite version 1.0.93.0 with dot >>net framework 4.0 and Visual Studio 2010. >> >> >>Obviously SQLite supports UNC paths because I am using SQLite database >>browser to open the same database using a UNC path. >> > > Also, a full pathname, can start with a double backslash (\\), indicating > the global root, followed by a server name and a share name to indicate > the path to a network file server. Just a little suggestion: UNC paths are slower than connecting that same path to a drive. If you are going to use it a lot, I suggest for you to connect that path to a drive and it will be much faster. We have a system using SQLite with a SharedDB and connecting that path to a drive is much faster. Ihth. josé -- Adolfo. > > ---- Mensaje original ---- > De: James K. Lowden > Para: SQLite mailing list <sqlite-users@mailinglists.sqlite.org> > Fecha: Mon, 06 Feb 2017 > Asunto: Re: [sqlite] Why does a query run 50x slower across a network? > >> In respect to the Windows environment I've appreciated that the use of UNC >> convention over a network (LAN) behaves much slower that "mapping" the >> logical unit as a drive letter D, E, .. Z in the local host. Altought >> unfortunately this doesn't seem very handy in all situations. >That's bizarre. By mapping a network file service to a drive letter, the user gains some convenience, and saves the OS very little: only the work of resolving the name, and maybe some other setup. Command conveyance and data transfer should be identical. In my experience, it always was. >If you're seeing noticeable difference, I'd expect you'll find they're either in name resolution or somewhere in the GUI. I can't think of any reason the underlying transport would be affected. _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users