Roman,

I was able to get around this problem by closing the connection
after periodic commits ( up to every 1500 records in parent table +
9000 records to the child table). This makes the database server
crash go away. I have the Java App that can be used as a test case
to recreate this problem consistently.I will attach the code and db
table structure first thing in the morning.

This side stepping makes me think that there seems to be an issue
with memory management in FB server.But I am also surprised that no
body has come across something like this.May be there is bug report
filed for this.

Thanks,
Karunakar Kona


--- In Firebird-Java@yahoogroups.com, "Roman Rokytskyy"
<rrokytskyy@...> wrote:

>
> > I am seeing the following msg in the FireBird log.
> >
> > -- terminated abnormally (4294967295)
>
> That's what I meant.
>
> > And I also came across a post by you related to this same msg
> > sometime in June 2004, that talks about Server crashing when too
> > many connections use up the available memory. The link for that

msg :

> >
> > http://groups.yahoo.com/group/Firebird-Java/message/5860
>
> The error code (4294967295) is -1 incorrectly interpreted. And -1

in this

> case does not mean anything that can say that issues are somehow

related.

>
> > I am researching to see if there is a relation between that issue
> > and my current problem. I am just creating one connection to the

db

> > and using it to do my inserts. I don't see any growth in the
> > transaction number on the database header page so I am assuming

that

> > I am not spawning multiple connections on db when the driver's

auto-

> > commit is turned off.
> >
> > Pls do let me know if there is something I am missing here.
>
> Best would be to create a reproducable test case in Java (most

likely that

> will require reproducing the same table structure, not sure about

data). In

> this case we can try to find the problem in the Firebird itself.
>
> Roman
>