Exadata Performance Diagnostics with AWR

” The contents of this paper apply to all Exadata deployments – whether onpremises, Public Cloud, or Cloud at Customer “

https://www.oracle.com/technetwork/database/exadata/exadata-awr-5100655.pdf

OPATCHAUTO-72043: Failed to create bundle patch object.

Problem:

Applying Patch Set Update (PSU) to Grid Infrastructure Home results in OPATCHAUTO-72043 error.

# /u01/app/18.3.0/grid/OPatch/opatchauto apply /u01/swtmp/28659165/ -nonrolling

OPatchauto session is initiated at Wed Feb 6 00:44:30 2019
System initialization log file is /u01/app/18.3.0/grid/cfgtoollogs/opatchautodb/systemconfig2019-02-06_12-44-32AM.log.

Session log file is /u01/app/18.3.0/grid/cfgtoollogs/opatchauto/opatchauto2019-02-06_12-49-13AM.log

The id for this session is WRHH
OPATCHAUTO-72043: Patch collection failed.
OPATCHAUTO-72043: Failed to create bundle patch object.
OPATCHAUTO-72043: Please verify the patch supplied.
OPatchAuto failed.

Solution:

Ensure unzip of the patch as grid home owner before running opatchauto from root user.

As a workaround, execute the below command as the <grid home owner> to fix the file permissions on unzipped patch location and re-run opatchauto from root user.

chmod -R 775 <unzipped patch location>

Check if a port on a remote system is reachable (without telnet)

Problem:

I wanted to check if our customer had an issue with network access. I asked him to run telnet command, but because of their security reasons, telnet rpm was not installed at all.

Solution:

One of the solution would be to run nc 🙂 but of course nc was not installed. 🙂

The following command helped me in that situation:

cat < /dev/tcp/192.168.3.4/3260

The port is open if there is no output, but if you receive the following, then the port is closed:

-bash: connect: Connection refused
-bash: /dev/tcp/192.168.3.4/3261: Connection refused

 

Steps that are necessary to run DBCA via Xming (Flashgrid-enabled clusters)

1. Remove immutable flag from /etc/ssh/sshd_config file:

# chattr -i /etc/ssh/sshd_config

2. Change the following in /etc/ssh/sshd_config:

From:

#X11UseLocalhost yes

To:

X11UseLocalhost no

4. Readd immutable flag to /etc/ssh/sshd_config file:

# chattr +i /etc/ssh/sshd_config

4. Restart sshd service:

# systemctl restart sshd

5. Start Xming on your computer and reconnect to the server via putty with X11 enabled.

6. Check current DISPLAY settings:

$ echo $DISPLAY
192.168.1.2:10.0

7. Copy Xauthority file from fg to oracle

$ sudo cp /home/fg/.Xauthority /home/oracle/.Xauthority
$ sudo chown oracle:oinstall /home/oracle/.Xauthority

8. Switch to oracle user and export DISPLAY variable, assign the same value that was returned from fg user (from where you run sudo su – oracle):

$ sudo su – oracle
$ export DISPLAY=192.168.1.2:10.0

9. Run DBCA:

$ dbca

GGSCI: error while loading shared libraries: libclntshcore.so.18.1: cannot open shared object file: No such file or directory

Problem:

Oracle GoldenGate Software Command Interface (GGSCI) 18c raises error:

$ /GG_HOME/home_1/ggsci
/GG_HOME/home_1/ggsci: error while loading shared libraries: libclntshcore.so.18.1: cannot open shared object file: No such file or directory

Cause: 

GoldenGate put all necessary shared objects (.so files) under its own home directory, but several shared objects may not be found there. They should be borrowed from RDBMS or GI home.

Solution:

# find / -name libclntshcore.so.18.1

/u01/app/18.3.0/grid/.patch_storage/28090523_Jul_14_2018_00_03_50/files/lib/libclntshcore.so.18.1
/u01/app/18.3.0/grid/.patch_storage/28655784_Oct_8_2018_21_27_28/files/lib/libclntshcore.so.18.1
/u01/app/18.3.0/grid/lib/libclntshcore.so.18.1
/u01/app/18.3.0/grid/inventory/Scripts/ext/lib/libclntshcore.so.18.1

Create symbolic link:

$ cd /GG_HOME/home_1
$ ln -s /u01/app/18.3.0/grid/lib/libclntshcore.so.18.1 libclntshcore.so.18.1

Rerun ggsci.

 

What is a Flex ASM and how to check if it is enabled?

In versions prior to 12c, the ASM instance needed to be run on each of the nodes of the cluster. In case ASM was not able to start, the database instance located on the same node was not able to come up also. There were a hard dependency between database and ASM instances.

With Oracle Flex ASM, databases are able to connect remote ASM using network connection(ASM network). In case of ASM instance fails, the database instance will reconnect to another ASM instance on another node. This feature is called Oracle Flex ASM.

Check if you are using such a great feature using the following command:

[grid@rac1 ~]$ asmcmd
ASMCMD> showclustermode
ASM cluster : Flex mode enabled

 

GGSCI: error while loading shared libraries: libclntsh.so.18.1: cannot open shared object file: No such file or directory

Problem:

Oracle GoldenGate Software Command Interface (GGSCI) 18c raises error:

$ /GG_HOME/home_1/ggsci
/GG_HOME/home_1/ggsci: error while loading shared libraries: libclntsh.so.18.1: cannot open shared object file: No such file or directory

Cause: 

GoldenGate put all necessary shared objects (.so files) under its own home directory, but several shared objects may not be found there. They should be borrowed from RDBMS or GI home.

Solution:

# find / -name libclntsh.so.18.1

/u01/app/18.3.0/grid/.patch_storage/28090523_Jul_14_2018_00_03_50/files/lib/libclntsh.so.18.1
/u01/app/18.3.0/grid/.patch_storage/28655784_Oct_8_2018_21_27_28/files/lib/libclntsh.so.18.1
/u01/app/18.3.0/grid/lib/libclntsh.so.18.1
/u01/app/18.3.0/grid/inventory/Scripts/ext/lib/libclntsh.so.18.1

Create symbolic link:

$ cd /GG_HOME/home_1
$ ln -s /u01/app/18.3.0/grid/lib/libclntsh.so.18.1 libclntsh.so.18.1

Rerun ggsci.