Saturday, 22 October 2011

Oracle Hyperion migration features

A nice amazement in the release is a novel value summoned hssmigrate. In beforehand liberations, a value summoned CSSImportExport did the Shared Services protection migration. Using this tool, you had to export protection from Shared Services, write an XML or CSV file
, update the file manually and afterward exercise the file to import protection into the aim environment. It sounds easy enough, but that the value is Java-based, which method it is case-sensitive.
Using expansion and demonstration settings as an case, users had to export protection from expansion, write the XML or CSV file, transfer it to demonstration, amend the file manually to reflect the goods and attendant labels and afterward import protection from the file. It was a manually intensive process.
Additionally, the case-sensitive aspect of the value had the probable to start headaches. Since there are no restrictions on naming conventions for external providers, population tend to call each one differently. There are no constraints in location to generate sure that the external provider call in the lineage setting (development) fits the aim setting (production). For a successful migration, the external provider labels had to match.

For Example
instance, declare you desired to migrate security from development to production. The provider label in Development is TOPDOWN (all caps). For a winning migration, you would have to recognise the provider label in development and, accordingly, bring ahead the matching provider label, employing the correct spelling and case, in production. The obstacle is that oftentimes more than one someone does the putting in of the development and goods produced environments. Therefore, the provider labels can end up out of sync. And at this purpose, you can't whole the migration. The only result is to correct the provider label in one of the environments.
Oracle acknowledged this obstacle and abolished it with the hssmigrate utility. The utility brings ahead an record on the source environment's Shared Services machine. You can then convey the record to the objective environment's Shared Services device, launch the configurator, pick Import Data from Earlier Release and you're done. The new utility doesn't care what the provider label is. It deals security to the surviving External Provider regardless. This collects many of time and vitality and abolishes the earlier subjects with server labels and wares suffixes, amid other things.
Another good new aspect is the Planning request for paid job migration wizard. The introductory steps to migrate a Planning request for paid job are still the same. However, the interface to perform the migration has changed. What employed to be a button on the Planning login piece of paper is now a nicer and more intuitive wizard on the Planning Administration page. The divergence is that you depart through Workspace to navigate to the Planning Administration page. Once on the Planning Administration piece of paper, you pick the Migration Wizard tab, where you can pick the request for paid job you like to migrate.
Everyone logs in through Workspace these days, so to depart through the Planning Web interface to find the Migrate button was not very intuitive. There is no new functionality with this process; in, you snap the Migration Wizard tab and then pick the request for paid job, and the wizard will redevelopment it. The new method is much more intuitive.

Oracle Hyperion migration drawbacks

What didn’t work so well in is the Essbase Studio migration. The putting in lead gave a step-by-step method, but it didn’t work, and my first endeavour at the migration was unsuccessful.
The migration of the Essbase requests for paid job in addition had a versioning issue. The conventional way to migrate Essbase requests for paid job is through Essbase Administration Services (EAS). It is as not hard as logging in, connecting to the source Essbase server and connecting to the objective Essbase server. You then launch the migration wizard, purpose to the source Essbase request for paid job, bring ahead the Essbase request for paid job on the objective Essbase server, chase a small number steps, and the request for paid job is migrated.
With, the method still works the matching way for Block Storage Option (BSO) cubes, but it flings an wrongdoing communication for Aggregate Storage Option (ASO) cubes. Behind the scenes, though, it becomes noticeable that the migration does carry the metadata over, and the ASO cube migrated successfully. This was substantiated by performing a post-migration high-level test. However, I have not wrapped up any low-level investigating, so I not able to chat to if the migration was complete. The other thing about the Essbase request for paid job migration is that the documented method is long and drawn out. After going through all the steps, you still end up with records that want to be conveyed from the source Essbase server to the objective Essbase server.

Final thoughts

Future Oracle Hyperion distributes should mend the Essbase Studio migration. That said, since it is a very new request for paid job, there are not many of population employing Essbase Studio. The earlier variety is buggy and the migration doesn't work, but bestowed the tiny customer foundation, fastening this is possibly not a main concern for Oracle at this time. The other good mend would be the ASO cube issue. Even with these two details, though, complete migration facilitates the hurt of the process.