transaction did not resolve till we bounced the Websphere app server.
And in that 30 minute interval, we got several error messages on the =
apps, forcing us to bounce the applications.
Once we bounced the application server, the transaction completed.
We are on 9.2.0.4 on Solaris 8.
thanks & regards
ratnesh=20
-----Original Message-----
From: oracle-l-bounce_at_freelists.org
On Behalf Of Powell, Mark D
Sent: Wednesday, May 05, 2004 3:14 PM
To: 'oracle-l_at_freelists.org'
Subject: RE: ORA-02056: 2PC state machine: bad two-phase command number
Based on the error information for the listed codes it would appear that
Oracle should clean this up given some time. Did Oracle manage to clean
this up or are repeated error messages showing in the alert log?
$ oerr ora 02054
02054, 00000, "transaction %s in-doubt"
// *Cause: network or remote failure in 2PC.
// *Action: Notify operations; DBs will automatically re-sync when the
// failure is repaired. Monitor pending_trans$ for final =
outcome.
$ oerr ora 02056
02056, 00000, "2PC: %s: bad two-phase command number %s from %s"
// *Cause: two-phase commit protocol error.
// *Action: recovery of transaction attempted. Monitor pending_trans$ =
table
to
// ensure correct resolution. Contact support.
If you have to clean this up manually then if you are using an undo
tablespace be aware of a bug in versions 9.0, 9.1, and 9.2 that results =
in
an ORA_30119 when you run dbms_transaction.PURGE_LOST_DB_ENTRY. If you =
hit
it then you will probably need to check metalink for the bug and fix: =
alter
session set "_smu_debug_mode" =3D 4;=20
HTH -- Mark D Powell --
-----Original Message-----
From: oracle-l-bounce_at_freelists.org
On Behalf Of Singh, Ratnesh (GEI,
GEFA, Contractor)
Sent: Wednesday, May 05, 2004 2:29 PM
To: oracle-l_at_freelists.org
Subject: ORA-02056: 2PC state machine: bad two-phase command number
Hi=3D20
this error happened on our prodn boxes, causing indoubt transactions.
Metalink does not seem to have much info on this.=3D20
It mentioned distributed_lock_timeout, which we have set at 180. =3D20
Does it need to be higher ?
Anybody seen this error, and knows what it is ?
Wed May 5 13:24:23 2004
Error 2056 trapped in 2PC on transaction 28.5.51628. Cleaning up.
Error stack returned to user:
ORA-02054: transaction 28.5.51628 in-doubt
ORA-02056: 2PC: 2PC state machine: bad two-phase command number 0
from =3D
prepared state(2)
Wed May 5 13:24:23 2004
DISTRIB TRAN =3D
1bba9.00000001C04EE8141C217F36843978DB9D51740BD3010ED5128B4F68
is local tran 28.5.51628 (hex=3D3D1c.05.c9ac)
insert pending prepared tran, scn=3D3D1064404425 (hex=3D3D0.3f7185c9)
thanks & regards
ratnesh=3D20
Please see the official ORACLE-L FAQ: http://www.orafaq.com
To unsubscribe send email to: oracle-l-request_at_freelists.org
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to: oracle-l-request_at_freelists.org
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to: oracle-l-request_at_freelists.org
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
