Whether you’ve been using Hyperion Financial Management (HFM) from its advent or you’re new to the Hyperion product, you’ve probably seen a lot of “best practices” guides. But it's just as important to know common “worst practices” that hurt HFM’s performance.
In my last blog post, I covered mistakes I frequently see organizations make with Hyperion Planning and Essbase. Now it’s time to talk about HFM.
Not tuning on a per app basis
Older versions of HFM required global performance tuning, but now most recent versions enable you to tune on a per app basis. Many customers have multiple HFM apps as well as historical apps, and tuning globally won’t give each app enough resources to function. When you tune on a per app basis, you can change the settings for each app — ensuring apps that do consolidating or calculating get more tuning than your historical applications.
Not changing expanision size for the database in SQL Server
If you’re using SQL Server, the default expansion size is one megabyte. When you set up your database and have one gigabyte of data, the SQL Server database will pause, add one megabyte, and resume. This will happen over and over and destroy your HFM performance.
Many tuning guides don’t mention having to expand SQL Server, and our consultants see this problem a lot. The solution is easy — set your SQL Server database to 1,500 megabytes so you always have some free space. If you’re using SQL Server and having HFM performance issues, this may solve all your problems.
Failing to check the logs
Your logs will be full of a lot of meaningless information, like people typing in their password incorrectly. But logs also contain meaningful information, like calculation times so you can see if they’re getting better or worse. Your logs might even reveal warning signs that you need to increase your connections or your max size. Checking the logs is a good way to find a problem before it becomes catastrophic, i.e. crashing during busy times.
Using Consolidate All
Using the Consolidate All with Data option is one of the worst things you can do in HFM. If you have empty cells, they will fill up with zeroes — meaning if your spreadsheet is 10 percent full, Consolidate All will make it 100 percent full. Now you’re calculating 10 times more data with no option to remove zeroes. To get rid of all the zeroes, you can do exports, delete your app, and rebuild it. The best option is to remove the Consolidate All with Data option in your security settings.