got ora-3113 while running pl/sql

[lgaorcl]

I got many trace file starting with: Got Ora-3113 while running PL/SQL.

Anybody else have had this problem?

Thanks!

[marist89]

Sure, you probably dumped a core file and a trace file. Review the trace file to find out what your problem really was.

[lgaorcl]

Yes, there are many trace files. But there are not so much useful message to me.

The following is a small part of it. Could you tell me anything from it?

Thanks!

OI***Got ORA-3113 while running PLSQL***
FUNCTION ODYSSEY.PS_GET_CONTROL_STRING:
library unit=30f80ad0 line=244 opcode=125 static link=0 scope=0
Current frame Registers:
FP=83f21c0 PC=30ef6844 Page=0 AP=8fabb40 ST=83f8f84
DL0=83f0780 GF=83f084c DL1=83f07e0 DPF=83f0844 DS=30ef7540
MCODE content in current Page:
Entry #0
ODYSSEY.PS_GET_CONTROL_STRING: 00000: BREAK
Entry #1
ODYSSEY.PS_GET_CONTROL_STRING: 00001: ENTER 28100, 0, 1, 1
ODYSSEY.PS_GET_CONTROL_STRING: 00009: INFR DS[8]+0
Frame Desc Version = 1, Size = 2163
# of locals = 200
TC_SSCALAR: FP+40, d=FP+6252, n=FP+6276
TC_ADT: FP+48, ltdo=HS+0, TRUE
_TC_iVCHAR: FP+64, d=FP+6280, n=FP+7792, mxl=1500, CS_IMPLICIT
_TC_iVCHAR: FP+84, d=FP+7796, n=FP+9808, mxl=2000, CS_IMPLICIT
TC_VCHAR: FP+104, d=FP+9812, n=FP+9816, mxl=32760, CS_IMPLICIT
TC_SSCALAR: FP+124, d=FP+9820, n=FP+9844
_TC_iVCHAR: FP+132, d=FP+9848, n=FP+10860, mxl=1000, CS_IMPLICIT
TC_SSCALAR: FP+152, d=FP+10864, n=FP+10888
TC_SSCALAR: FP+160, d=FP+10892, n=FP+10916
TC_SSCALAR: FP+168, d=FP+10920, n=FP+10944
TC_SSCALAR: FP+176, d=FP+10948, n=FP+10972
TC_SSCALAR: FP+184, d=FP+10976, n=FP+11000
TC_SSCALAR: FP+192, d=FP+11004, n=FP+11028
_TC_iVCHAR: FP+200, d=FP+11032, n=FP+11052, mxl=11, CS_IMPLICIT

[julian]

3113 is extremely hard to trace. Could be anything. Even if you open an iTar with Oracle, you might not get the answer to "Why?". But I suggest you open the iTar and upload the trace files and let them think :-)

[veeramur]

ORA-3113 "end-of-file on communication channel" is a catch-all error message which simply means that the foreground oracle client process (svrmgrl, sqlplus, RMAN, Oracle Forms etc.) has lost contact with its background shadow process.

Unfortunately there are many causes of this. Network glitches were clearly the cause that the Oracle developers had in mind when they phrased the
error message, but anything which causes the shadow process to crash will cause this error to be reported.

This message may also be accompanied by other messages, either displayed on the screen or in log files. If the background process which has crashed
is one of the essential Oracle servers processes (PMON, SMON, DBWR etc.) you may also get an ORA-1034 ORACLE not available error
message. Looking in the alert log or looking through any trace files in your background_dump_dest may help diagnose the problem.

Common causes are:

Incompatible values of the init.ora parameter processes and the kernel shared memory and semaphore parameters

This is a very frequent cause because the default installation of Oracle sets up incompatible values!

Oracle executables need relinking

shared_pool_size too small.

timed_statistics=true.

This is probably version and path-level specific. But at least one version of Oracle benefits drom switching off timed_statistics.

a datafile extended over 2Gb in size.

Create table as select .... where there was an invalid procedure or function called from a trigger defined on the table we were selecting from.

Lot of time I have received this error,some times the network problem,some times shared pool size proble.So please check your environment everything is ok.

Thanks
Veera