The problem concerns distributed transactions.There are 2 databases that communicate with each other through a database link.We call Xa transactions from our weblogic application. Both databases work in a shared server mode.
Everything works until we run some performance tests on the evironment where the data volume is much bigger then the one on the other environments where the same tests succeeded. We're getting the following error: ORA-03113: end-of-file on communication channel
The error get registered in a local database alert.log when accessing data through the database link while there's no trace corresonding to this error as far as the second database is concerned. We're almost sure there's no network problem. The error occurs in only 1% of a total executions number. We also observe the inactive sessions number increasing on the remote database.
Both databases use ASMM by having set the SGA_TARGET and PGA_AGGREGATE_TARGET parameters and proper level of the STATISTICS_LEVEL parameter, so some hot elements like SORT type operations should execute fine.Any Oracle parameter should be tunned or we should take into consideration something else...
I start sql*plus and start working. after 2-3 minutes i get error -
ORA-03113: end-of-file on communication channel.
I googled and found to see alert.log.(but i found no such file on my system) but in my system i found some info at C:\app
ishiproduct11.1.0db_1log ishi-pcclientclsc42.txt
the content of the file is ;
******************************************** Oracle Database 11g CRS Release 11.1.0.6.0 - Production Copyright 1996, 2007 Oracle. All rights reserved. 2011-01-27 12:35:03.512: [ OCROSD][3332]utgdv:1:could not open registry key SOFTWAREOracleocr os error The system could not find the environment option that was entered. 2011-01-27 12:35:03.512: [ OCRRAW][3332]proprinit: Could not open raw device 2011-01-27 12:35:03.512: [ default][3332]a_init:7!: Backend init unsuccessful : [33] 2011-01-27 12:35:04.617: [ CSSCLNT][3332]clsssinit: error(32 PROC-32: Cluster Ready Services on the local node is not running Messaging error [9]) in OCR initialization *************************************************
I am getting the same error evert time i login.Tried to stop/start oracle services but problem persists...
We are running Oracle 11.1.0.6 on CentOS Linux at our local network. Two dozen users work from client's network to access this database. There are firewall and switches between the local network and client network. Everything run smoothly for last two years until about 10 days ago, something changed. The client on the client's network can still connect to this database, but they can not run large queries. Their client connection hangs and time out. I can see TNS-12535: TNS:operation timed out and TNS-00505: Operation timed out in the alertlog file. They even had problem to load stored procedures and views onto TOAD from your PC. All the user on the local network can work on this database normally.
We consulted network team and firewall team, none of them said they make any changes recently and have no firewall rule or others to block the traffic.
I have a strong feel that this is a network related problem. But don't know how to provide solid evidence to make network people/firewall people to buy it.
You can connect to the database but can't run query on large table/view?
We had power failure, and i had to restore the database from offline backup.after restore i tried to start the database and i received following errors.
SQL> startup ORA-32004: obsolete or deprecated parameter(s) specified for RDBMS instance ORACLE instance started. Total System Global Area 663961600 bytes Fixed Size 2184056 bytes Variable Size 394267784 bytes Database Buffers 260046848 bytes Redo Buffers 7462912 bytes Database mounted. ORA-01589: must use RESETLOGS or NORESETLOGS option for database open
[code]....
I reviewed the /oracle/PRD/saptrace/diag/rdbms/prd/PRD/incident/incdir_583361/PRD_ora_7179_i583361.trc, it hugh file with alot of Jargon.
I made the number of processes to 1 by issuing the below command.
SQL> alter system set processes = 1; SQL>shutdown immediateI am getting the below error when I issued startup command.
SQL> startup; ORA-03113: end-of-file on communication channelalert log: Fri Dec 28 14:09:24 2012 PMON started with pid=2, OS id=2984 Fri Dec 28 14:09:24 2012 PSP0 started with pid=3, OS id=5788 Fri Dec 28 14:09:25 2012 ORA-00020: No more process state objects available ORA-20 errors will not be written to the alert log for the next minute. Please look at trace files to see all the ORA-20 errors. Process GEN0 submission failed with error = 20 USER (ospid: 4780): terminating the instance due to error 444 Fri Dec 28 14:09:25 2012 VKTM started with pid=4, OS id=1676 at elevated priority
Instance terminated by USER, pid = 4780How to startup my database again?
I have an app that collects data and after some time submits the data to Oracle.
The app. has worked flawlessly for years in different setups.
Now, I have a setup where the clients are in a separate LAN behind a NAT router and the server is on the main LAN.
I get "ORA-03113: end-of-file on communication channel" in different circumstances.
Sometimes it happens, when the app has not been in contact with the database for 15-20 minutes. This can probably be handled by disconnect/connect mechanisms.
But whats worse, sometimes ORA-03113 appears in the middle of continuously submitting hundreds/thousands of rows.
The server does not have routes to access the clients separate LAN, is that part of the reason?
I have two machines with oracle 11g . Both having ODP 11.2 Beta. I've to communicate between these two machines using Oracle Advanced queuing. The message sent from one machine will reach the second machine only if the Firewall of both machine is OFF.But based on requirement, we cannot switch OFF firewall or add exceptions as ports or applications.
One of the customer is using Oracle database Oracle database 9i and its listener is running on port 1521. When Oracle database client are connected to database , it is connecting on multiple ports ( like 1600 to 1690 ) . if it is possible to restrict Oracle database client connection to only one specified port or not .
Is it possible to communicate with a serial device (via COM1) with oracle forms / webutil? I am able to communicate in java using RXTXcomm, but I am having problems making a bean that uses it (getting security errors and yes, I have signed the jars).
Basically, I have a scale hooked up to my COM1 port and want to be able to send commands (zero and tare) to it and read the weight from it through a form or java bean.
When executing a SELECT with JOIN in oracle error as I am having the title. If I change the JOIN to where the procedure works normally. When installing my application, it works normally for a day or two, then it only shows the end of communication error. What I do is select the following:
I got a primary database which ships logs successfully to a standby database where it is applied successfully as well.
I can connect using the tns name defined for data guard (which are already used for log shipping) in sqlplus.
Yet, at the time of configuring data guard broker,the command add database 'standby' connect identifier is 'to_standby'; fails.
Looking into drc<instance>.log, I see the error ora-03113 network I/O.
I haven't seen any error like that online and I can't explain why it does that when logs can ship already with the same connection identifier.
I have edited dest_2 with and without parameters LGWR and SYNC and reconfigured the broker without success. I've removed the configuration files and tried again.
I've checked that the password file is the same on both servers as well.Nothing seems to work. Is there a parameter I'm missing somewhere.
I have a table that contains a CLOB column with pseudo-XML in it. I want to keep this data in an XMLType column so that I can leverage some of Oracle's built-in XML features to parse it more easily.
The source table is defined as: CREATE TABLE "TSS_SRM_CBEBRE_LOGS_V" ( "INCIDENT_ID" NUMBER, "EVENT_TYPE" VARCHAR2(100 BYTE) NOT NULL ENABLE, "EVENT_KEY" VARCHAR2(100 BYTE), "CREATION_DATE" TIMESTAMP (6) NOT NULL ENABLE, "CREATED_BY" VARCHAR2(100 BYTE) NOT NULL ENABLE, "LOG_MSG" CLOB);
The target (for testing this problem) table is defined as: CREATE TABLE "TESTME" ( "LOG_MSG" "XMLTYPE") My query is: insert /*+ APPEND */ into testme ("LOG_MSG")select XMLTYPE.createXML("LOG_MSG") as LOG_MSG from "TSS_SRM_CBEBRE_LOGS_V" b; In SQL*Developer, my error is: Error report:SQL Error: No more data to read from socket In SQL*PLUS and Toad, my error is: ORA-03113: end-of-file on communication channelProcess ID: 13903Session ID: 414 Serial number: 32739
it is necessary to allocate channel for rman back up and during rman back up what should be the state of target database me script for configuration and incremental level 1 daily back up script
and normal oracle secure incremental level bacl up script back up should be done daily
whats the difference between channel and parallelism in RMAN ?
What is the Oracle recommendation for RMAN channel allocation ? I have read some were that RMAN channel allocation must be 4 times of the Server CPU -1. Say for example if there are 2 cpu in the server, then the channel allocation would be 4 * 2 -1 = 7 ?
what is the meaning of 'SBT_TAPE' in the following command in RMAN? In the following command, does it refer to a tape drive, or a hard disk backup?
allocate channel t1 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=<a backup server server name>,NSR_CLIENT=<a oracle server>,NSR_DATA_VOLUME_POOL=ORACLE ARCHIVE LOG)';
Backup for database has been working fine until today. Today, about 1/3 of the way through the backup (large database), I got this... RMAN-12005: error during channel cleanup RMAN-10038: database session for channel sbt_4 terminated unexpectedly ...and the backup died.
Never seen before and the research I've done online has been less than useful. This database is an SAP/oracle db with brtools 7.20 (22) and is using the new parameter in its init.sap file, called "_rman_chan_failo = yes", to address disagreement between brtools and oracle about channel drops being ok if all datafiles backed up.
Nothing in the alert log at the time of this error.
while executing a clone script i got the below error : -
RMAN-00571: =========================================================== RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS =============== RMAN-00571: =========================================================== RMAN-03002: failure of restore command at 11/15/2011 20:07:34 RMAN-12010: automatic channel allocation initialization failed RMAN-06758: DB_UNIQUE_NAME is not unique in the recovery catalog
Here i understand that RMAN couldn't find which SPFILE to use to restore the target DB.
How do i check for the DB_UNIQUE_NAME in recovery catalog?
The proposed solution that i found was Use the FOR DB_UNIQUE_NAME option to specify the name of the instance whose parameter file you want to restore. how to implement this in my clone script?