One common reason your dashboards may not be working as intended is that your ElastiCube builds are not designed properly. This isn’t a catch-all problem, and though there are multiple reasons why it can occur, the outcome is the same: insights aren’t accurate, dashboards can’t access data, or information is displayed improperly.
The issue with build failures is that they aren’t always readily visible and discovering them requires understanding underlying causes for errors. Fortunately, these errors are usually not catastrophic, and each have straightforward fixes.
[H2] The Top 4 Reasons ElastiCubes Fail, and How to Solve Them
These are some of the most common reasons your ElastiCubes can malfunction and the best ways to quickly resolve problems as they arise:
- Data Modeling Issues - Many-to-Many Relationships
Smart Data Modeling is the backbone of a healthy deployment but some common mistakes, such as Many-to-Many relationships, can undermine your analysis and put unnecessary load on your Elasticube performance. A Many-to-Many relationship exists when the value in each field used to create a relationship between tables is included multiple times in each table.
Many-to-Many relationships can cause duplications in the returned datasets, which can result in incorrect results and might consume excessive computing resources.
The Solution:
- Test your schema to see if it includes many-to-many relationships
- Understand which scenario best fits your current schema
- According to your schema logic, applying the respective solution
- Check out these assets to dig deeper:
- Connectivity Issues – Problems Communicating
This occurs during the initial setup and can lead to accessibility issues as well as problems gathering data. Connectivity issues can be triggered by anything from incorrect credentials, improperly set up connections and servers, and other similar errors.
The Solution: Check for different connectivity issues and make sure your ElastiCubes are properly configured. Another area to investigate is your database availability, for example scheduled maintenance or database restrictions (too many database connections). Check out DbVisualizer, a database tool that let’s you explore your data sets and database as related to the ElastiCube.
- Not Enough Disk Space – Memory Issues
Disk space problems come in two varieties: a lack of storage space, or the more common issue of insufficient RAM to process the existing ElastiCube. The latter is a common error when your ElastiCube schema include too many columns, tables, and connections, and can cause your system to crash. You can notice this during the creation process if there are disk space errors.
The Solution: Start by trimming your ElastiCube schema by removing tables, data, and fields that are not essential. Removing unwanted data can show a quick turnaround. Otherwise, you can add more RAM to your machine to handle the larger load.
- Testing ElastiCube Models on Live Servers – Performance Problems
It’s difficult to know exactly how an ElastiCube will perform before deploying it, as errors may not be apparent until it’s interacting with other programs. Deploying the ElastiCube on a live server, then, is an unknown element that can have adverse effects on the larger system. These can include improper functions, broken systems, and in some cases fatal loss of data.
The Solution: Before placing anything on your live server, you should always test it on a development server to avoid breaking anything before you know how to resolve it. Testing on a dev server also gives you the chance to test for unexpected errors and make tweaks before uploading an ElastiCube to the live servers.
What To Do If Your ElastiCube Fails
When you notice a failure in your builds, it’s not always easy to identify the culprit. If you do notice an issue, try the following:
- Explore Sisense’s Build Errors list to review some of the most common errors for ElastiCube builds. The chart features known errors and their standard solutions.
- Read through the Sisense documentation for strategies and tips on how to troubleshoot build errors.
- If you can’t find a solution on your own, submit a support ticket to Sisense’s staff.
- Speak directly with your Customer Success Manager, who will be able to give you guidance on how to resolve your issues.
