Issues/Symptoms |
Causes to Review |
Details |
General Story Slowness |
Too many widgets in the story |
The IDP defines the ideal weight-based approach (tested by SAC) to the number of widgets one should have in the story to avoid. Refer to the IDP or SAC Best Practice Document defining the weight-based approach for the number of widgets |
|
Bloated Data Model |
The data model defined in the query designer should not be bloated or have an unnecessary number of widgets. The report designer should add only columns and object schemas on the need basis and functional design of the story report |
|
Too Many Data Sources |
Some report designers try to bypass the suggested 120-column limit in the data source by introducing multiple data sources and linking them which in turn can lead to further performance deterioration |
|
Too Much Data Blending |
While there are scenarios that require data blending, some report designers try to put so much information into a single page leading to unnecessary data blending which can cause problems |
|
Nesting too many Calculated Columns |
Nesting calculated columns can lead to performance issues as building calculated columns over other calculated columns mean sequential processing leading to a longer time to render. In some scenarios wherever feasible, it is better to build calculated columns directly over source columns rather than building intermediate calculated columns |
|
Too many Self Joins |
Some reports require multiple joins to the same object in the query designer. Some report designers use this as a mechanism as a prerequisite to show the transposition of values at the widget in the story (row level value now being column headers).
This can be achieved by using the cross tab at the story level and still having single schema objects at the query designer layer
We will be adding a detailed blog on the same |
|
Wrong Use Case- Data Dump and SQL Analytics |
Some use cases like dumping data to excel or complex window function-based SQL Aggregates are best suited for Workforce Analytics or other SAP tools like SAC Planning. Consult with your CSM and Partner on the same
For the partial export use case of Table Widget data export (for limited data volume), you can refer to the following blog here. [Note: Report Stories is not a Data Extraction Tool]
Alternatively, there is also a scheduling facility for similar export use cases. More details to follow |
Slowness in working with filters |
Improper Use of Filters |
There are multiple scenarios with improper use of filters that can cause issues. A few of them to review are as below. We will define some of them below
1) Not using scope filters for providing matrix level filters i.e manager level/ HR level data access
2) Not using schema object level filters before joining 2 schema objects. This is critical to avoid data explosion at the time of data fetched for the query
3) Not using filters on story widgets
4) Using the same filters at multiple places
5) Not using bookmarks features where one needs to work a group of filters with specific values and store those as bookmarks
We will add other blogs on the Filter usage soon |
|
Apply Selection for Filters |
Not using the “Apply selection for filters” feature in a Story. This one is a recently supported feature for facilitating multiple value selection in high cardinality filters.
Refer to the blog for more details
|
|
High Cardinality Filters |
High Cardinality Filters take a long time to load as they have a lot of values to load at the time of selection. This can be optimized by reducing some of the not needed values by filtering them at the query level only |
|
Avoid linked Dimension on Calculated Dimension and then adding the same as filters |
Using a Calculated Column as a Linked dimension can have a performance impact |
Story Timeouts and Errors |
Scope Filtering |
While a lot of these errors were addressed and resolved in b2205 and b2211, if such issues are still encountered, the best way is to apply proper scope filtering to reduce the amount of data fetched to prevent timeout |
|
Missing RBP permissions |
Another reason for the widget-specific errors could be the story viewer not having RBP permissions for the specific columns used in the widgets |
|
Authorization Errors |
Sometimes users are not synced properly due to IDP configuration issues. Do refer to the IDP on the same |
Wrong data or No Data in the widgets |
MDF Objects unavailable for Reporting |
MDF object reporting needs to be reviewed for proper configuration. Please refer following help section on custom MDF object reporting limitations.
Refer to the blog for more details |
Table Widget slowness/ failed to load/ widget crash |
Drill Limitation Removed and 1 million cell limit breach |
Currently, there is a limit of 1 million cells for the table widget and in case the drill limitation is set to unlimited and the result set has more than 1 million cells, the query guardrail would kick in and the widget would fail
The best would be to split the table into multiple tables with row-level filters
For a few scenarios, offline scheduling can be considered where the report is a table widget alone |
|
Building Complex Hierarchies via Calculated Columns |
Hierarchies are currently not supported in Report Stories. Bypassing the same by building these hierarchies via calculated columns can impact performance and should be avoided
We will be sharing a detailed blog on the topic soon |
Chart Widget Slowness |
Top N filter on high cardinality dimensions |
Chart widgets using high cardinality dimensions can be slow to run, using the "top N" filter can help improve performance |
Imported Stories Fail to Run |
Content Store Import failure |
The following blog provides a comprehensive guide to fixing errors in import.
At the same time do refer to the Query Validation tool for changes in the imported query vs the actual query [Refer to the Blog here for QVT] |
Raising Support Tickets for Customer Support to investigate |
Attach HTTP Archive (HAR) files to debug for Engineering review |
For the support team to investigate customer-reported issues, it is best to provide HTTP Archive (HAR) files as attachments while raising a ticket.
Refer to the blog for more details |
