How big can a Cube be?

We have a number of growing projects in one cube. My customer want to know what is too big for a data cube in eazyBI. Any guidance on when a cube is too big?

1 Like

Hi @jeffscudder

Thank you for the valuable question.
There are no specific technical limitations on how many issues are ok for eazyBI: we have seen accounts with less than a thousand issues with very slow reports and accounts with 5oo thousand issues running normally.
The performance depends more on how the reports are created, what information is imported, and how frequently imports are run.

  1. Reports are optimized.

In large accounts, reports should be high-level (by projects, years), and preferably should contain only predefined or JavaScript-precalculated measures.
For analyzing data in more detail (user or issue level reports, daily/weekly reports, history analysis, etc), separate smaller accounts might be created.

Here are overall best practices on how to create optimized reports: Best practices for report creation

To monitor the health of reports, use usage statistics to see which reports are slow; they will tend to get even slower when the projects are growing, therefore consider optimizing them. Contact eazyBi support for guidance with that!

  1. Import only information you need

In the account, import only the projects that need to be analyzed together. For project-specific reports, create separate accounts for each project.
You may also apply JQL filters to exclude a specific set of issues from importing and analyzing (for instance, issues closed long ago): Data from Jira
Select only those custom fields you need. You can always come back to import options and add more fields.

  1. Import frequency

Import data from Jira less often (preferably once a day), especially for larger accounts. During data import, the cache is cleaned, and reports are opened faster if they are already cached. If you run imports every 10 minutes, then the cache is cleaned all the time, and reports are always opened from fresh, which is slower.

Additionally, you may watch a couple of presentations how to plan and maintain accounts: