Thursday, 30 November 2017

...Restatements



One of the topics that I have spent time on is to understand the most efficient way to restate reported data in HFM.

On one side, there are multiple applications which either do not handle this requirement or the process to maintain and analyse the data is too complicated.

  • The most usual case is to restate the P12 data. To handle this requirement, there are applications with an additional period P13 to handle the late adjustments. I think that this approach has been inherited by the design of the most ERPs.
  •  Another case is to restate the data from all the periods and not only the P12 data in case of a change in the reporting standards or an audit. In this case, there are applications that have a specific scenario which is called RestatementX.
  •  In a similar case of restatement, in order to restate the data of all periods, the application had two sets of accounts. The accounts with the reported data and the accounts with the restated data. The restated accounts had the accounts with the reported data as share members in addition to the unique restating accounts.
  • Finally, there are applications that do not maintain the reported data and before posting the adjustments, the users must run and hardcode all the values in order to keep the reported figures (bad, bad, bad)


On the other side, HFM is a really powerful and flexible tool and we should use it in order to transform the focus of our processes from data intensive to data analysis.

This means that we should be able to restate our numbers without much effort, without having to copy data from one scenario to another or one period to another and at the same time allow us to audit the numbers and trace the flow of the change. The easiest way to implement such change is to introduce a new dimension in the HFM application (HFM version 11 has unlimited number of dimensions) which will allow us to introduce versioning. 

As an example, let’s say that you will use Custom5 as your version dimension. Custom5 will have 3 members: [None], Restatement and AllCustom5. The [None] member will be linked with Phase 1 and the Restatement member will be linked with Phase 2. The process is simple. The users will load all the initial data in the member [None] and you will lock Phase 1. These data are the reported data. When you will need to Restate your numbers, the users will load any adjustments in the member Restatement and you will lock the Phase 2. 

In this way, when you will retrieve data from [None], you will read the reported data and when you will retrieve data from AllCustom5, you will read the restated data.

Disclaimer: The above example is really simple and will not cover all the cases. Allow your phantasy to work and adapt this mechanism in your requirements.

Thursday, 16 November 2017

...Smartview and OneDrive



Recently (July 2017), I discovered a new functionality of OneDrive which allows multiple users to work on the same file concurrently and I think it is called real time collaboration. This seems to be an amazing feature if you are preparing a presentation and you are collaborating with someone who is in the other side of the world. Unfortunately, it can be a huge inconvenience if you are using SmartView.

Thursday, 9 November 2017

...Financial reports and Shared Services



A few months ago, we were ready for our monthly submission. Everything was quiet and I was relaxed enjoying our Cloud environment...until we received an alert from our monitoring tool that the first of the two FR WebLogic servers auto-restarted. Of course, having a WebLogic server auto-restarting was not a problem since that was the process. Until five minutes later

Thursday, 2 November 2017

...CPU utilisation in HFM



There are 3 settings that manage the CPU utilisation.

  •  MaxNumConcurrentConsolidations: The number of concurrent consolidations that can run in a server at one time. This setting will adjust the total number of consolidations that run in one server regardless the number.
  • NumConsolidationThreads: The number of CPU treads that can be used to multi-thread a consolidation. A consolidation with 32 threads will finish quicker than a consolidation with 8 threads but the first one will use all the available CPU.
  • ConsolidationMultiThreadingScheme

Thursday, 26 October 2017

...Memory utilisation in HFM


Since version 11.x, Oracle has released the 64-bit version of HFM. The benefit is that the 64-bit version of HFM can hold more data in memory comparing to the 32-bit version.

Based on Oracle documentation, HFM virtual address space is dictated by the OS. So, if you have an application with large subcubes (large scenarios, dense data, big number of scenario) which struggles in the 32-bit environment,

Thursday, 19 October 2017

...Records & Subcubes in HFM




A few years ago, I decided to spend time to understand how HFM calculation engine works with aim to improve the performance of the rules. The first stop was the HFM manual, the basic design considerations and the section that is referring to subcubes to understand the calculation mechanism.

Wednesday, 19 July 2017

...writing!



Hi all,


Just to say that I am still around since a friend told me yesterday that I do not write any more.


Actually, I was thinking that I have not posted anything for a few months and I need to post one of the topics that challenges me but I have just been tied up with the reviewing our processes and preparing application updates.


The good news is that as soon as I will finish with this I will participate in another amazing project the coming year which hopefully will introduce a new way of reporting data.


Blog soon.


Thanos

...Restatements

One of the topics that I have spent time on is to understand the most efficient way to restate reported data in HFM. On one side, ...