Subject | Re: Slow connection..It's a bug...... |
---|---|
Author | alexp772001 |
Post date | 2002-06-11T19:54:03Z |
I noticed this problem on the following operating systems:
- Win 2000 Professional
- Win 2000 Server.
- Win 2000 Advanced Server.
- Windows XP Professional (.gdb renamed into .fdb)
on Intel and AMD processors (only tested on monoprocessor machine).
Linux is not affected by this problem.
The first connection is immediate, the following have to wait about 1-
3 seconds.
Regards,
Alessandro Petrelli.
- Win 2000 Professional
- Win 2000 Server.
- Win 2000 Advanced Server.
- Windows XP Professional (.gdb renamed into .fdb)
on Intel and AMD processors (only tested on monoprocessor machine).
Linux is not affected by this problem.
The first connection is immediate, the following have to wait about 1-
3 seconds.
Regards,
Alessandro Petrelli.
--- In ib-support@y..., Brad Pepers <brad@l...> wrote:
> > I don't believe this "Slow connection...It's a bug..." thing.
I've
> > **never** observed it, not in IB 5.6, not in IB 6.x, nor in any
of the
> > Firebird builds. The issue I'm aware of with the IB local client
on
> > Windows is that it's not thread-safe.
> >
> > Using the following setups I have sub-second connection speeds in
all cases
> > with Firebird 1.0. The Windows server is NT SP 6a running on a
slowish
> > Celeron (434 MHz, 384Mb), the Linux server is RedHat 7.2 on a
850Mhz AMD
> > Duron, 512 Mb.
>
> You don't see it with your particular setup and therefore you don't
believe
> everyone else that does??? Come on!
>
> I too have seen this problem but it only effect my Windows XP
system. I have
> put it down to the whole .gdb filename suffix thing. People keep
saying to
> rename the database to some other suffix (I'm using .fdb) but won't
the
> isc4.gdb file still be a problem? I don't know of a way to change
its name
> and every time a user logs in I would think the server is accessing
that
> database for user info and having the .gdb problem.
>
> Change your server to Windows XP and then tell me you don't see the
problem.
>
> --
> Brad Pepers
> brad@l...