Shulou(Shulou.com)05/31 Report--

This article mainly explains "how to patch PSU", the content of the article is simple and clear, easy to learn and understand, the following please follow the editor's ideas slowly in depth, together to study and learn "how to patch PSU" bar!

First, let's talk about how to download patches and the latest opatch:

All on mos:

Download the patch: search directly for 2118136.2, which is the document id

Download optach: click the Patch and Update tab and search for 6880880

Next, let's talk about the commonly used PSU:

We are talking about linux here, mainly DB,GI+DB,GI+DB+OJVM.

Then let's talk about patch installation:

In fact, it is here to translate readme.html, and here we mainly introduce GI+DB (+ OJVM).

OJVM just remember that you need to use oracle alone to play, and this does not support rolling upgrade, that is, the cluster must be stopped.

Then say GI+DB:

First of all, no matter what you say, check the version of opatch first.

Both grid users and oracle users should check under their respective $ORACLE_HOME/OPatch/:

. / opatch version

Then use this version to compare what version of opatch is required for the patch in readme. If you ask me, it's right to upgrade the latest opatch, the way to upgrade opatch:

Download the corresponding system version of OPatch, and then execute it under both grid and oracle users, because each has its own Opatch tool:

Unzip package-d $ORACLE_HOME

Then choose to overwrite all, that is, enter an A to ok.

Then to generate a rsp file, anyway, it will be generated under grid: (12.2.0.1.5 or later is not needed)

$ORACLE_HOME/OPatch/ocm/bin/emocmrsp-no_banner-output / opt/ocm.rsp

Anyway, no matter where the rsp is put, it must be accessible to both oracle users and grid users. Later, why?

Then everything is ready, and root creates an empty directory, which must be an empty directory, such as:

Mkdir / GI

Then copy the package to this directory and give it 777 permissions:

Chmod 777-R / GI

Then use the grid user to extract the patch package and extract it under this path:

Su-grid

Cd / GI

Unzip Patch Pack

Generally speaking, neither grid home nor oracle home belongs to share. If there is an acfs, you have to umount first, and then we will call directly with opatch auto. Here we use rolling upgrade, which is a node-by-node fight without stopping business.

According to the documentation, you have to stop the em first.

Then execute as root user:

Grid's $ORACLE_HOME/OPatch/opatch auto / GI/GI+DB directory-ocmrf / opt/ocm.rsp

When finished, under grid users and oracle users, execute:

$ORACLE_HOME/OPatch/opatch lspatches

See if the patch has been patched, and then all the nodes are done.

Then execute the script on one node:

@? / rdbms/amdin/catbundle.sql psu apply

@? / rdbms/admin/utlrp.sql (this is mainly to recompile some java-related invalid objects) (if you have time to execute, even if you don't have time)

Then check dba_registry_history, there is a new patch record is right, done.

Now explain the opatch auto in readme:

First of all, if you do not build that empty directory, the system will ask you whether you unzipped to an empty directory, ah, you confirm twice, because it is to retrieve each directory itself, and then patch it accordingly.

Then, it stops the em, the database instance, the asm instance, and the node

Next, it patches GI with oh=grid_home specified by the user of grid, and patches DB with oh=oracle_home specified by oracle user, where the script for cluster operation is executed as root, so that's why:

Be sure to patch with root execution! Be sure to put the ocm.rsp file in a path that both grid and oracle can access! (oracle users cannot access / home/grid/)

After the patch, the cluster is up, the database is up, and so is em.

So to put it bluntly, we should do it in the way of readme, there is no need to stop anything, just acfs umount it.

However, in order to ensure insurance and avoid many unforeseen problems, it is recommended that you do the following:

Stop em and database instance on the node to be patched

Sqlplus / as sysdba

Shutdown immediate

Exit

Grid's $ORACLE_HOME/OPatch/opatch auto / GI/GI+DB directory-oh grid_home 's path expansion (meaning don't write $ORACLE_HOME directly)-ocmrf / opt/ocm.rsp

Oracle's $ORACLE_HOME/OPatch/opatch auto / GI/GI+DB directory-oh oracle_home 's path expansion (meaning don't write $ORACLE_HOME directly)-ocmrf / opt/ocm.rsp

Then lspatches checks

Then start the database instance, and then stop the database instance at the next node

All nodes are patched, and then a script is executed on one node to register the patch in the database.

Thank you for your reading, the above is the content of "how to patch PSU", after the study of this article, I believe you have a deeper understanding of the problem of how to patch PSU, and the specific use needs to be verified in practice. Here is, the editor will push for you more related knowledge points of the article, welcome to follow!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.