Subject | Re: Type 2 driver modifications are ready for review. |
---|---|
Author | ryanb486 |
Post date | 2003-05-31T15:54:12Z |
--- In Firebird-Java@yahoogroups.com, "robert_difalco"
<rdifalco@t...> wrote:
make
use of a more efficent local ipc method(although currently at least
for
windows the implementation has drawbacks too) or in the case of the
embeded
server - no ipc method at all. This can under some circumastances(eg
running
many individual but small fast querys) lead to almost doubling of
performance.
I found when doing some benchmarking that type 2 mode when making a
tcp/ip
connection to the server is actually slower then in type 4 mode
overall.
<rdifalco@t...> wrote:
> Anyone know why there is a need for a Type 2 driver?The primary benifit comes when the type 2 mode allows the java app to
make
use of a more efficent local ipc method(although currently at least
for
windows the implementation has drawbacks too) or in the case of the
embeded
server - no ipc method at all. This can under some circumastances(eg
running
many individual but small fast querys) lead to almost doubling of
performance.
I found when doing some benchmarking that type 2 mode when making a
tcp/ip
connection to the server is actually slower then in type 4 mode
overall.
>are
> R.
>
> --- In Firebird-Java@yahoogroups.com, "Robert DiFalco"
> <rdifalco@t...> wrote:
> > Out of curiosity, what is the value of the Type 2 driver? What
> the
> > main advantages of the Type 4 driver already present with a new
> Type 2
> > driver? Are there benchmarks that illustrate the difference in
> > performance?
> >
> >