By Daniel Mortimer-Oracle on Mar 03, 2015
What is Zero Downtime Patching?
With Zero Downtime Patching (aka ZDT), a patch can be rolled through an Enterprise / High Availability implementation, one component at a time. This ensures that the system can remain available and functioning during the patching process. Many Fusion Middleware products now support this type of patch. However, note the applicable versions below:
- 11g - patchset 184.108.40.206 or higher must be in place first
- 12c - 12.1.2, 12.1.3 or higher
A ZDT patch allows the patched and non-patched components of the system to run side-by-side to provide continuous service.
In the diagram below, a 2-node cluster is running two Managed Servers. The Managed Servers are part of the same Cluster. This means they are running the same applications and any server can fulfill a user request.
Each node or machine also has a separate Oracle Home. In this example, a ZDT patch is applied to one Home at a time.
The starting state:
1. Server on Node1 - server1 - is shutdown
2. Patch is applied to the Oracle Home from which server1 is running
3. Restart server1
At this point, the instance of App A on server1 is patched (running from a patched Oracle Home) but the instance on Node2 - server2 - server2 is not patched.
For a ZDT patch (and a ZDT patch only) this state is acceptable during the patch window.
Finally, the patch is rolled through the rest of the system:
1. Server on Node2 - server2 - is shutdown
2. Patch is applied to the Oracle Home from which server2 is running
3. Restart server2
The patch has now been applied to the entire system. At no point was the entire system down.
How to identify a ZDT patch?
A Zero Downtime Patch is identifiable by the instructions in the patch's README file.
The README should say, for example:
This patch has been marked as eligible for Zero Downtime Patching. The type of Zero Downtime Patching supported by this patch is FMW_ROLLING_ORACLE_HOME.
If this phrase does not appear in the patch README then this is not a Zero Downtime Patch and should be applied as normal. following the instructions in the Patch README.
This blog entry is intended to give a heads up to "zero downtime patching." If you are interested in finding more it is important to read support document:
Zero Downtime Patching in Fusion Middleware (Doc ID 1942159.1)
which provides detail about:
- ZDT patching prerequisites and limitations
- In-place and Out-Of-Place patching
- Other uptime options which you may see in a patch readme