ImpactEPM Recovery

The ImpactEPM Recovery solution provides online backups in the EPM instance for instant recovery as well as offsite backups for your Planning and FCCS implementations.

Oracle "It is your responsibility to archive backups for any duration your organization requires". The ImpactEPM Recovery can bridge the gaps in your organizations EPM Backup and Disaster Recovery strategy.

How to recover data video

Overview Self Service Recovery Disaster Recovery Synchronise & Encryption Development Assets Back Story

Overview

ImpactEPM Recovery Overview

1 - By keeping data backups online, this enables power users to invoke a restore of data to a known point in time

- In the TEST instance, backups of the artefacts provides security of the development assets.

2 - Porting copies of the data and artefacts to a second data centre enables Disaster Recovery to be covered.

3 - Offsite storage for longer retention of archives.

The basic offering directly on the EPM platform has only one nightly backup available, anything else will require additional management. The Impact EPM Data Recovery service offers far more security and flexibility.

	Impact EPM	Oracle Snapshot
RPO (Recovery Point Objective)	Nightly or more often	Nightly
Data History Retention	31 days plus 11 month end	Previous night only
RTO (Recovery Time Objective)	Business to run on demand	Administrator to Create whole instance
Backup Files Support Life	12 Months+	One Monthly patch cycle
Multiple Copies of Development Backup	Yes	No
Multiple Copies of Meta Data Backup	Yes	No
Off Site Storage	Yes	No
Business Self Service Recovery	Yes	No
GDPR Compliant	Yes	Yes
Backup transferred to second Data Centre for DR	Yes	No

SaaS Industry Practice

"A snapshot is compatible with the previous and next updates of an environment. For example, a snapshot from the 19.09 update of Oracle Enterprise Performance Management Cloud will be compatible with 19.08 and 19.10 environments, and can be imported into these environments."

Generally these backups are designed to be used by the provider to ensure it meets its customers’ SLAs with regards to uptime and high availability. In case of a major outage, the SaaS provider has backup processes in place to protect its data, systems, and applications.

Jira, Shopify, Confluence and GitHub are SaaS providers that have 3rd party services for backup.

Self Service Recovery

The backup of data to the same instance enables Power Users to initiate a data recovery without the intervention of IT support with Administrator Privileges. A 10 day cycle of files is kept available in the instance, older files can be recovered from the Off Site server.

One of the key features of the ImpactEPM Recovery is that the data of an individual database in an application can be restored, rather than having to restore the whole application.

For Business Power Users

Directly within the application the power User can initiate the recovery, the EPM process sets up the data load jobs ready for use.

From the iEPM Task List, select Import and then select the data to import for the model to be recovered.

For Administrators

Data will be stored around a 10 day cycle in the inbox/outbox explorer found in “Overview”

Import of the data is possible by a Service Administrator using the data import jobs that the service creates

Disaster Recovery

Impact EPM Backup Service Details

With the second SaaS datacentre, here the main models are in London and the secondary is Cardiff, this backup process will copy data and artefacts there and load them overnight to form part of an Enterprise wide DR solution.

If the second datacentre instance was in use for another Business Process, the backups would only be copied and not restored to the instance.

Due to the nature of the data export, it is much smaller than the default Artefact Snapshot, reducing significantly any likelihood that the Oracle data limits will be exceeded. The Oracle snapshots can be very large, easily exceeding 150Gb and not being able to be stored on the second data centre instance

EPM Cloud Backups for OCI Gen 2 (oracle.com)

Synchronise & Encryption

Synchronize Data

In both Production and Test environments the metadata, calculations and other artefacts are backed up.

ImpactEPM recovery, uses a logical backup like the Artefact Snapshot enabling full application restore to different instances, or update of individual components. It differs in approach in that it separates programs from data. This enables the following benefits:-

No expiry of the Snapshots after one full update cycle (about 6 weeks).
Data can be restored without having to recreate the whole instance.
Data and Dimensions can be synchronised from the Production to the Test environment.

Synchronising will simplify the reconciliation process of new developments to the Production environment.

Encryption in Transit

Communications between ImpactEPM server and EPM instances uses HTTPS, the default that is used by EPM Automate and all user connections to EPM.

Encryption at Rest

“If your organization's policy allows you to encrypt content at rest with an Azure-managed key, then no action is needed - the content is encrypted by default. For managed disks, the content inside storage is encrypted by default with Server-side encryption with platform-managed key. The key is managed by the Azure Storage service.”

FAQ - Azure Disk Encryption for Windows VMs - Azure Virtual Machines | Microsoft Learn

Development Assets

Development Assets Secured

In both prod and test environments the meta data, calculations and other artefacts are backed up.

Securing the development assets for project work keeping recovery points for developers to revert to previous versions of code.

All the artefacts are stored in the “Migration” utility in EPM. It will cycle around an optional 10 days copies of the artefacts.

Each snapshot contains all the Artefacts for the environment.

Back Story

Why Recovery

Innovation driven by necessity

The use case started on a programme with a major Whisky producer, the situation the project encountered was the need to get to a restore point, but not really knowing which day to recover from. Valuable time was spent in trial and error to identify the correct backup to get to a known reconciled point in the budget lifecycle.

Having this solution it would have taken minutes to see each restore point rather than the hours it did take, thus it would've had the budget process back on track the same morning rather than the next day.

How it would have helped in a practical sense is that copies of the Production back up data are stored on the PROD system and also the TEST system, facilitating immediate restore of the data only for the validate of the budget.

With this, on a later project at an insurance firm there was a requirement as part of the project to create from scratch a system to backup four environments (DEV, TEST, PREPROD, PROD). Additionally it had to move the backups between the environments, and then set up a process for updating artefacts into the production environment as part of the change control process.

Another project that would have benefit from this in a very different way. A financial services company that had been acquired were changing their year end date. Trying to reconcile the TEST environment to the PRODUCTION environment which updated every day, a considerable amount of time was spent extracting data, loading to TEST and reconciling. The automation from Impact EPM Recovery would have made this an automated quick and reliable process, cutting a lot of manual setup work.