| Title: | DB Integrator Public Conference |
| Notice: | Database Integration - today! Kit/Doc info see note 36 |
| Moderator: | BROKE::ABUGOV |
| Created: | Mon Sep 21 1992 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 1171 |
| Total number of notes: | 5187 |
Hi,
I have a customer that have a problem with Transparent gateway
for Rdb with oracle7. He is getting a error 1PC_UPDATE. He wrote a program
that in first transaction he has read one oracle7 table and wrote on Rdb
and he started a second transaction to delete the table on oracle7 and
he got the error above.
I've tried to find out the meaning of this error in ldrv_msg.doc,
but I've failed.
Please, Someone could tell me where I could get this meaning or the
meaning of this error (sorry, but I don't know dbi enough).
Thanks...Pedro
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1137.1 | ldrv_msg.doc isn't helpful, sorry... | BROKE::ABUGOV | Thu Mar 27 1997 14:41 | 37 | |
Hi,
From the product family users guide:
2.3.3 Specifying the Transaction Mode
The gateway products cannot fully participate in
distributed transactions that require the coordinated
commitment of update operations to multiple databases.
Although the gateway products do not prohibit participation
in a distributed (two-phase commit) transaction, they
do not support them; they only support one-phase commit
transactions. You will get an RDB-E-1PC_UPDATE error
message when you attempt to write to a gateway database
in a distributed transaction. You can disable this error
by using the TX_MODE=NOWARN_1PC qualifier, as shown in the
following example:
SQL> ATTACH 'FILENAME/TYPE=SYBASE/USER=JIM-
cont> /TX_MODE=NOWARN_1PC';
To enable the warning message, use the following syntax:
SQL> ATTACH 'FILENAME/TYPE=SYBASE/USER=JIM/TX_MODE=WARN_1PC';
The TX_MODE qualifier is not required. The default is for
the warning message to be enabled.
Basically, the message is to let users know explicitly that there is
not a 2 pc transaction going on, and their heterogeneous/distributed
databases can end up out of sync when they update. To allow the update
to happen, add the tx_mode=nowarn_1pc.
Hope this helps,
Dan
| |||||