Subject | Re: [IBO] OT: FW: SuperIB Event Alerter Announcement |
---|---|
Author | lester@lsces.globalnet.co.uk |
Post date | 2001-01-31T08:13:20Z |
Svein
I did not mention it, because I thought that closing the table (
xx_query/xx_cursor ) did close the transaction. Is this possibly where I
am going wrong?
Certainly when I close the connection, everthing is 'disconnected', but
events need the connection open. Where I am actually copying
information, there is a Start/Commit, but is it possible for a
transaction to be active if all the table type links are closed?
I do not rely on default transaction, so only my 'real' ones should
exist.
Svein Erling Tysvær wrote:
I did not mention it, because I thought that closing the table (
xx_query/xx_cursor ) did close the transaction. Is this possibly where I
am going wrong?
Certainly when I close the connection, everthing is 'disconnected', but
events need the connection open. Where I am actually copying
information, there is a Start/Commit, but is it possible for a
transaction to be active if all the table type links are closed?
I do not rely on default transaction, so only my 'real' ones should
exist.
Svein Erling Tysvær wrote:
>
> Lester,
> closing tables/queries doesn't allow for the OAT to advance, but committing
> (using Commit, not CommitRetaining) transactions does. You probably do this
> already, but I thought I should mention it since your mail didn't mention
> transactions at all.
>
> Set
>
> At 07:23 31.01.2001 +0000, you wrote:
> >The only thing that I would add is although events are nice, I still
> >have 'growth' if I leave their connection permanently open. I have to
> >drop and remake every hour to ensure that OAT can roll on preperly. This
> >may be me, but MOST of the time, all tables and other components have
> >been closed and are not opened until the event announcing a change is
> >recieved, which can be several hours at a time.