> From: trafalgar42 [mailto:trafalgar42@...]

> I run InterBase as our security database which receives large volumes
> of hits daily. When I was running the 1.0 SuperServer, the
> main database
> engine would eventually exit (after maybe 2 months of up-time) with no
> explination. As a result, I've had to switch to Classic
> Server which solves the
> main problem. However, I really like the way that the SS
> caches data for
> all clients. Furthermore, InterBase documentation indicates
> that the
> SuperServer architecture is the future of InterBase - I
> don't won't to have to operate on a depricated system.
>
> Has anyone out there experiences such a problem? If so, is there a
> patch available or what have you done to solve this issue?

In one of the recent changes done for Firebird v1.5, a change was
committed to solve a bug whereby the server would crash under heavy
load. I can't say that this is your problem.

As to the future of Interbase, I can't speak to that, but there has been
some interesting discussions in the Architect mailing list arguing that
using Classic as the base would provide a more logical path for
Firebird. One of the main reasons is that Classic doesn't suffer from
the known SMP problems and that resolving the obvious
contention/sharing/caching/locking issues of the Classic server would be
easier to achieve than trying to resolve the SMP problems of the SS
implementation.


Sean