| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> Re: Distributed transactions (two-phase commint) without DB links
> Well, in our case this IS exactly the problem.
> Developer has to know
> where he does what.
For me insert into tab2_at_remote is good enough.
Developers will be more happy to do just that instead
of doing XA transactions.
insert into tab1;
insert into tab2_at_remote;
commit;
Now, if somebody can put the XA API code instead of this. Not sure what developers will prefer?
> What about delayed flight? What about 100?
For me the same. Your system will die from time to
time.
If some flight transaction is not finished or half way
finished you can recover from that stage. You should
always know who is the master db and recover it.
Of course my points may be stupid if several parties are involved (several companies, products, ...).
I am just trying to see the big hint reverting me from discussion here :)
> If XA API calls are so expensive - sure it's not the
> way to go. In
> fact we are still using Tuxedo (6.5) and moving away
> from it. I
> thought if we can use from the client the same
> mechanism as db links
> but under our control from OCI than it would be good
> alternative.
In OCI you can do everything. Even Oracle is not
telling us directly that PL/SQL is based on it, it is
on some lower level for a sure.
Do not believe that ORacle maintaining 3 or four
different layers of interaction with the core db.
OCI is great. you have even direct path API,
callbacks, things that PL/SQL was still not using or
did not use in earlier releases.
Of course, it is hard to me to judge what is the best in 10 minutes reading our posts.
Regards,
Zoran Martic
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Mar 14 2005 - 18:17:37 CST
 |
 |