Re: timing problem

Tony Leake wrote: On Thu, 2005-09-29 at 16:15 +0100, Nuno Pereira wrote: [EMAIL PROTECTED] wrote: This one is interesting in terms of concurrency... Is the app2 the one responsable for setting the flag? I supose that it is. If that happens it's important that app2 doesn't mess with the fla

Re: timing problem

Nuno Pereira <[EMAIL PROTECTED]> wrote on 09/29/2005 11:15:08 AM: > [EMAIL PROTECTED] wrote: > > Tony Leake <[EMAIL PROTECTED]> wrote on 09/29/2005 07:08:24 AM: > > > > > >>Hi > >> > >>I have 2 applications communicating via a mysql database > >> > >>the db is 4.1.8 running on a debian linux sys

Re: timing problem

On Thu, 2005-09-29 at 16:15 +0100, Nuno Pereira wrote: > [EMAIL PROTECTED] wrote: > > This one is interesting in terms of concurrency... > > Is the app2 the one responsable for setting the flag? I supose that it is. > If that happens it's important that app2 doesn't mess with the flag, i > mean

Re: timing problem

[EMAIL PROTECTED] wrote: Tony Leake <[EMAIL PROTECTED]> wrote on 09/29/2005 07:08:24 AM: Hi I have 2 applications communicating via a mysql database the db is 4.1.8 running on a debian linux system. All tables are innodb app 1 1 runs on a windows machine, is written in c# and talks to the

Re: timing problem

On Thu, 2005-09-29 at 09:30 -0400, [EMAIL PROTECTED] wrote: > The key here is > that you are making multiple changes from app1 that really should be > within a transaction. The entire process of writing sales data and > unsetting a flag from app1 needs to be transacted. That way the other >

Re: timing problem

Tony Leake <[EMAIL PROTECTED]> wrote on 09/29/2005 07:08:24 AM: > Hi > > I have 2 applications communicating via a mysql database > > the db is 4.1.8 running on a debian linux system. > All tables are innodb > > app 1 1 runs on a windows machine, is written in c# and talks to the db > with odb