Hi Elan,
When you want to use "remote" you have to have the server running.
When you want to use "local" you cannot have the server running.
Only one process can use the "local" connection to the database at a
time.
Basically, the server uses local access to the database, and provides
the option of any* number of "remote" access connections.
https://groups.google.com/group/orient-database/browse_thread/thread/5eb45a3a47a9977f/4dd5842fcd4d9a66
has some additional information about the difference.
There are also some differences when using the graph mode, e. g. when
using TinkerPop. For example, transactions aren't supported for
remote: mode access yet.
I can't easily find the threads discussing that or other differences
here, but check out
https://groups.google.com/group/orient-database/browse_thread/thread/7c738f5b2c20cc00/f2291dd02367a474
as one example.
Hope that helps,
WBT
(*) "Any number of" is likely not accurate. There's an upper limit
somewhere, but that doesn't appear relevant here. Luca will correct
me if I'm wrong. :-)
> "local" use intra-JVM to connect to the server. The JVM process opens the
> files without the need of a server. For the "remote" is more like classic
> RDBMS where a server listen for incoming remote connections.
>
> Lvc@
>