Re: In a HA environment: Do triggers fire when a database record gets updated?

[CRPence <CRPbottle@xxxxxxxxx>]

  Although it is correct to suggest that the database does not care 
what application performs the actual I\O request corresponding to the 
trigger type, the trigger will always fire if enabled and operative. 
Note: That excludes the /system/ as the application, because IIRC, the 
system interface can establish an effective override to suppress firing 
a trigger; e.g. given the implementation of the online reorganize 
feature, one must presume such a feature is utilized, because concurrent 
update activity can verify that application I/O still experiences the 
effects of triggers while the RGZPFM activity is unaffected.

  However applying of journal changes, which implements some HA, are 
not actual I\O requests [none of the I/O methods are utilized] and thus 
that activity will never fire a trigger.  I am not sure of the effects 
for using the Replay Database Operation (QDBREPLAY) API; an HA provider 
using that feature, might choose the option to disable triggers for any 
trigger additions being replayed.?

  So... it really matters how the HA is implemented and possibly even 
how the HA is setup to perform [possibly using one of different 
available implementations to apply changes], whether the triggers will 
fire on the target system due to their means to apply the changes at the 
target.

Regards, Chuck

On 15 Apr 2013 11:01, Charles Wilt wrote:
> My understand is that the answer is yes, triggers are a function of
> the DB and the DB doesn't care who initiated the change. <<SNIP>>
>
> On Mon, Apr 15, 2013 at 1:51 PM, David Gibbs wrote:
>
> > I've got a customer who has a HA environment (I'm not sure what HA
> > system though).<<SNIP>>
> >
> > Do database triggers fire, on the slave system, when a record
> > update is replicated from the host?  Is this dependent on the HA
> > system?