Category | Best Practices Recommended | Roles/Persona |
Editable App
|
Always create templates for your critical editable apps to avoid losing them. |
Authors
|
Harvesting feature must be used only when necessary, and it should be replaced with data retention outside Klera. | ||
Regularly clean up/purge unnecessary data in the editable app. | ||
Remove redundant formulas on a regular basis. | ||
Remove editable app from schedules if not required. | ||
Optimize execution flow of editable app periodically using the new features/functionalities added in Klera. | ||
Keep containers in "Re-use mode= never" unless data is being harvested. | ||
Do not use editable app Copy option just for versioning. It is best to use templates for versioning. | ||
Archive the unused/stale editable app on a separate disk and delete them from Klera. They can be re-imported when needed. | ||
Ensure Klera version compatibility while importing editable app on your Klera instance. | ||
Templates
|
Always keep the output of the Root Operation on the first page of the editable app, which is to be saved as a template. |
Authors
|
Unused/redundant templates should be removed regularly from the system. | ||
When creating form-based templates, the output container of the form should be kept on the same page as the form. | ||
Periodic audits must be done to optimize the templates with new visualizations, computation methods, etc. | ||
Make sure none of the containers is on Replace/Append mode unnecessarily. | ||
Make sure that the tools/systems used in the template are pre-configured for the user who is going to use the imported/published template. | ||
To use any Master data CSVs in a template, ensure it is imported outside the editable app where the template is being created. This will ensure that the template does not prompt its user to reimport the CSV at the time of execution. | ||
Always design a template for accessing relevant data only. Reduce the dataset at each step to relevant records before executing the next operation, using pre-filters. | ||
Before saving the final template, ensure the removal/purging of unnecessary/redundant datasets and formulas. | ||
Publishing
|
Before publishing, always hide editable app pages/tabs which are not intended for the users. |
Authors
|
Verify the published link before you broadcast it to the intended audience. | ||
Republish the editable app to reflect the changes made after it was published. | ||
For scheduled editable app with auto-publishing enabled, always monitor publishing status through alerts. | ||
Ensure Klera version compatibility while publishing to a remote Klera instance. | ||
Schedulers
|
Ensure your schedules are staggered well across the day and avoid overlaps. |
Authors
|
Manage schedules centrally through a power user (administrator/ Author), to avoid redundancies, resource crunch, and overlaps. | ||
Ensure that the schedule frequency is set considering the time taken for completion. Do consider a minimal gap between two schedules for handling overruns. | ||
Filters & Formulas
|
If a formula takes longer than 5 minutes to complete, consider re-writing the logic or reducing the input dataset. |
Authors
|
Aggregation/Table Scan type of formulas are compute-intensive; try to create them optimally and remove such unnecessary formulas. | ||
While doing a Full Join or Cartesian Join, be cautious of the data size of both the datasets as these joins produce large datasets. | ||
Operations
|
If an operation takes longer than 5 minutes to complete, consider re-factoring/ splitting the input request. |
Authors
|
While executing data access operations, ensure that the query scope is optimal and only relevant data is fetched. | ||
Try to avoid running new operations if previous operations are still in the running state. | ||
Connectors
|
Users must always use instances of a specific connector and not the connector itself. |
Connectors
|
Always upgrade the instance, if the connector is upgraded/modified. | ||
Always keep a backup of the connector before upgrading/modifying it. |