Z
ZaX
Good day folks,
I'm working with a third-party transactionnal SDK, packaged into a few DLL.
In a particular scenario, I need to have two sessions opened on the
infrastructure I access thru the SDK, and I need programmatical control over
the two sessions, so I can make changes in one session depending on the
information I receive from the second session and so on.
However, according to the third-party SDK support staff, a flaw in their
architecture prevents me from running two sessions on a single instance of
their assembly... So they suggested me to run two instances of my program
instead. This works.
But the code need to be extremely performant, so using System.Remoting to
communicate between the two instances of an application would be the least
acceptable option.
So my idea was to try to link to two different instances of the assembly in
some way... Any idea on how I would do that? Should I use system.reflection
to instantiate my types?
Other suggestions, comments?
Thanks!
ZaX
I'm working with a third-party transactionnal SDK, packaged into a few DLL.
In a particular scenario, I need to have two sessions opened on the
infrastructure I access thru the SDK, and I need programmatical control over
the two sessions, so I can make changes in one session depending on the
information I receive from the second session and so on.
However, according to the third-party SDK support staff, a flaw in their
architecture prevents me from running two sessions on a single instance of
their assembly... So they suggested me to run two instances of my program
instead. This works.
But the code need to be extremely performant, so using System.Remoting to
communicate between the two instances of an application would be the least
acceptable option.
So my idea was to try to link to two different instances of the assembly in
some way... Any idea on how I would do that? Should I use system.reflection
to instantiate my types?
Other suggestions, comments?
Thanks!
ZaX