It may be that you need to set your TIB_Transaction.TimeoutProps such that
you get a guaranteed closure of transactions after a specified period of
time.

I think it is possible I give people the idea they don't have to even think
about OAT issues with IBO. But, IBO in most cases handles it automatically
but there are necessary precautions that need to be taken because the only
parts of which IBO automates OAT advancement is that which it can do
transparently. All the rest of the cases, programmers are expected to define
how they want their behaviors to be carried out.

In invite all who are using IBO and have not made a specific effort to
understand transaction OAT advancement issues to come to the IBO public
support group and get the whole scoop on this critical topic. There has been
a lot written on this subject as well.

Regards,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com


""Luciano Sparacino"" <lsparacino@...> wrote in message
news:00d901c1a35e$f75fd270$6d0001c5@......

> I'm getting a similar error with IB6 and FB RC1 but i always get the

server to slow down performance at a specified time everyday (always close
to noon). Once the server is restarted, the application starts working ok
again.

>
> Our application also uses IBO and TIBOTable and TIBOQuery components to

access the database and most of them uses autocommit. We are actually
changing the code to manage explicit transactions (with
TIB_TransactionSinlge) for the main operations and we started to get better
results (at least the server recovers after a user finishes to ask for a
heavy query). I don't know how IBO manages Autocommit, but it's not a
subject to be discused in this group.

>
> When we examine the IB log we find the message " gds__free: attempt to

release bad block " like 10 times per second before it gets slow and needs
to be restarted.

>