Recently I had issue with extensive trace generation on new Exadata system. There was no ORA- errors in alert log and adrci show incidend command returns no rows but the common line in almost every trace was IPCLW line. tail alert_GDPRD5U4.log ARC1 (PID:111419): Archived Log entry 22126 added for T-4.S-5539 ID 0x1b27ae48 LAD:1 2018-10-01T13:34:54.117829+00:00 Thread 4 advanced to log sequence 5541 (LGWR switch)   Current log# 41 seq# 5541 mem# 0: +RECOC1/GDPRD5U/ONLINELOG/group_41.26674.978267741 2018-10-01T13:34:54.262530+00:00 ARC2 (PID:111447): Archived Log entry 22130 added for T-4.S-5540 ..read more

This entry is based on MOS: How To Relink The Oracle Grid Infrastructure Standalone (Restart) Installation Or Oracle Grid Infrastructure RAC/Cluster Installation (11.2 or 12c). (Doc ID 1536057.1) In case OS Linux update, we need to think about relinking Oracle Software. This post show how to perform relinking in Oracle Clusterware(18.4) configuration after OS update from OEL 7.6 to 7.7: [oracle@rac02 ~]$ cat /etc/oracle-release  Oracle Linux Server release 7.6 [oracle@rac02 ~]$ uname -a Linux rac02 4.14.35-1844.0.7.el7uek.x86_64 #2 SMP Wed Dec 12 19:48:02 PST 2018 x86_64 x86_64 x86_64 GNU ..read more

One of a new feature of 18c is refreshable pdb switchover which is documented under: https://docs.oracle.com/en/database/oracle/oracle-database/18/multi/administering-pdbs-with-sql-plus.html#GUID-B505C234-FAF4-4BAB-8B59-59276E0EA128 I performed a couple of tests and want to present the biggest disadvantage of this feature: during switchover, the source pdb database is fully recreated. This can be unacceptable for large databases as the switchover process will take a long time depending on you IO and network bandwidth. For sure this unpleasant side effect can limit use cases. To show you this ..read more

Based on metalink note(Measuring Network Capacity using oratcptest (Doc ID 2064368.1)) I've checked the network performance for one Oracle DataGuard enviroment. Here is the approach: [oracle@marmi101 ~]$ ping marmidr PING marmidr.cmd.far.xtr.com (10.28.32.151) 56(84) bytes of data. 64 bytes from marmidr.cmd.far.xtr.com (10.28.32.151): icmp_seq=1 ttl=64 time=0.458 ms 64 bytes from marmidr.cmd.far.xtr.com (10.28.32.151): icmp_seq=2 ttl=64 time=0.317 ms --- marmidr.cmd.far.xtr.com ping statistics --- ---lvm /srv/ora/app/oracle/12.1.0.2/jdk/bin/java -jar oratcptest.jar -help java -jar oratcptest ..read more

In this post I would like to shows you the validate function implemented in Oracle 18c DataGuard Broker. The test enviroment consist of two Exadata X7-8 with Oracle Grid Infrastructure and Oracle Database 18c software. Primary and Standby are RAC 2 nodes databases. Check the configuration(just to have an overview): DGMGRL for Linux: Release 18.0.0.0.0 - Production on Thu Jul 5 11:32:52 2018 Version 18.1.0.0.0 Copyright (c) 1982, 2018, Oracle and/or its affiliates.  All rights reserved. Welcome to DGMGRL, type "help" for information. Connected to "testdb" Connected as SYSDBA. DGMGRL> sho ..read more

The SYSAUX Tablespace can get full due to many reasons but in my case it was due SM/OPSTATS segments(WRI$_OPTSTAT_HISTHEAD_HISTORY and their indexes) which occupied most of the space. How to fix the issue and avoid adding more and more space to the tablespace?. This post shows the root cause and how to repair it and it's based on real production datawarehouse(30TB) running on Exadata X4-2. DB Version is 11.2.0.4 BP20. Let start with standard query checking the "occupants" in SYSAUX tablespace: set linesize 120 set pagesize 100 COLUMN "Item" FORMAT A25 COLUMN "Space Used (GB)" FORMAT 999.99 C ..read more

Duplicate with noopen clause is very useful in scenario when you perform migration(to other server) and upgrade(to new release) on new server where old Oracle Software is not installed. In my testcase I used NFS server to save the backup from source database. It was presented to new server. Just for your info: I used it to upgrade from Oracle 11.2.0.2 version on Red Hat 6.9 to Oracle 12.1.0.2 on Red Hat 7.4. Please check MOS for more info: Upgrade to 12c(12.1 /12.2) through RMAN Duplicate using BACKUP LOCATION with NOOPEN clause (Doc ID 2022820.1) So let's start with preparation. On source d ..read more

With end of 2015 will Oracle Database 11g out of normal support so lot of customers make an upgrade campaign in data centers. Every DBA go(should go at least) through a list of manual step to upgrade database (Complete Checklist for Manual Upgrades to Oracle Database 12c Release 1 (12.1) (Doc ID 1503653.1)). This note state that after upgrade we have to set new value for COMPATIBLE parameter. I will recommend as well to look on another instance parameter related with database release software version - OPTIMIZER_FEATURES_ENABLE. This parameter is responsible for enabling a series of optimizer ..read more

Number of failed login attempts exceeds threshold value message comes from OEM. How to check what user is causing this? DBA_AUDIT_SESSION can help to identifiy the username,userhost, etc: SQL> desc DBA_AUDIT_SESSION  Name                -------------------  OS_USERNAME         USERNAME            USERHOST            TERMINAL            TIMESTAMP           ACTION_NAME         LOGOFF_TIME         LOGOFF_LREAD        LOGOFF_PREAD        LOGOFF_LWRITE       LOGOFF_DLOCK        SESSIONID           RETURNCODE          CLIENT_ID           SESSION_CPU         EXTENDED_TIMESTAMP  PR ..read more

Using oracle function dbms_workload_repository.AWR_REPORT_TEXT with spool and linux comand grep, you can analyze load statistics or even track performance peeks without continuously access to your production database. I know that now we have more and more applications which can deliver such information(for example AWR Datawarehouse). But if you are command line geek… Check this outJ All we need is dbid, instance id, begin_snap and end_snap_id to generate following string: SELECT output FROM TABLE(dbms_workload_repository.AWR_REPORT_TEXT(:dbid,:instid,:begin_snap,:end_snap)) 　 I create ..read more