LCM and High Availability
High availability A system attribute that enables an application to continue to
provide services
in the presence of failures. This is achieved through removal
of single points of failure, with
fault-tolerant hardware, as well as server
clusters; if one server fails, processing requests are
routed to another
server.When our Shared Services is set up for high availability and is started
as a Windows service we need to configure LCM as well for high availability.
If our shared services is on two servers then of-course each
server will have its own
import_export folder from where respective Shared
Services will pic the file system backup
information, so we need to have a
shared disk or shared location which can be used by both
servers when we
export artifacts using Lifecycle Management. In order to do it we configure
LCM
so that the contents are exported to a path on a shared disk, when imported,
the content
is read from the
shared disk exported location.
To configure Lifecycle Management for high availability:
1 Set up a shared disk/folder that is accessible to all Shared Services nodes. (Here, for ease I
1 Set up a shared disk/folder that is accessible to all Shared Services nodes. (Here, for ease I
have granted permission to everyone but you can restrict it
to the Domain User who will be
starting services )
2 On each node, start Shared Services as a service using the login of a domain user who has
access to
the shared disk/folder.
3 On one node, launch Oracle's Hyperion® Shared Services Console and expand the
Deployment
Metadata node under the Foundation application group.
4 Expand the Shared Services Registry node, then Foundation Services, and then
Shared
Services.
5 Under the Shared Services node, right-click the Properties node and select Export for Edit.
6 Save the component.properties file to a location on the file system.
7 Open the saved file in a text editor and search for the property
filesystem.artifact.path.
8 Change the value associated with the filesystem.artifact.path property.
UNIX-style UNC paths with forward slashes must be defined for the shared disk;
for
example:
Configuring Lifecycle Management for Shared Services High Availability 25
filesystem.artifact.path=//hostname/share
The empty part is the host name or IP address.
9 Save the changes.
10 From Oracle's Hyperion® Shared Services Console, right-click the Properties
node under
Shared
Services, and select Import after Edit.
11 Browse to the location of the updated file and select the file.
This action updates the property in Oracle's Hyperion Shared Services Registry.
12 Restart Shared Services on this node and all other nodes using the domain
user login.
Little Details:
Lets assume that after configuring you decided to take the
File System backup for one of your
applications then there is might be a chance
that you end up with below error. You must ensure
that you have provided full
control to the user who will be starting services to the shared folder;
for
most of the security related issue you will end up in the below error:
EPMLCM-37077: Cannot
perform migration. Read/Write permission to the repository path -
//<HostName>/SharedImportExport/admin@Native Directory/hj is not
available. Give
permission and rerun the migration.
If you are on 11.1.2.0 then the Patch Set Exception (PSE)
11696634 has to be applied
to enable using UNC paths.
In UNIX,
inMIDDLEWARE_HOME/user_projects/epmsystem1/config/FoundationServic
es.
2. Set the value of filesystem.artifact.path
as follows:
filesystem.artifact.path=MOUNTED_FILE_SYSTEM_PATH
In this example, replace MOUNTED_FILE_SYSTEM_PATH with the actual path of
the mounted
shared disk; for example, /net/myServer.
3. Save and close the migration.properties file.
4. Repeat the preceding
steps on each node in the cluster.
Cheers...!!!
Comments
Post a Comment