This page helps you choose your Fredhopper deployment model.
Concurrent vs. Cascading data publishing
Based on our experience, we have identified two major deployment models that have emerged in eCommerce. The main difference between these models is driven by the catalogue or data publishing process.
- Concurrent data publishing: In this model, the PIM or Data Management Application is concurrently and automatically publishing data into both the Live environment and the Preview environment.
- Cascading data publishing: In this model, data is first published into the Preview environment, before being published into the Live environment. Please see Cascading Deployment Model for more details on this deployment model.
Choose your deployment model
This section provides a detailed comparison of the two deployment models, comparing general characteristics, platform characteristics and the supported Fredhopper business processes.
General comparison
The table below provides a general comparison of the deployment models.
| Characteristics | Concurrent data publishing | Cascading data publishing |
|---|---|---|
| Industries (typical) | Travel Publishing/Classifieds Price comparison Entertainment business |
Multi-channel retail Mail order |
| Platform (typical...) | Custom platforms | Standard platforms (Hybris, IBM, Intershop, MS) |
| Data flow | Single-stage process: Concurrently from PIM to both Preview and Live |
Two-stage process: From PIM to Preview, and from Preview to Live |
| Data update process | No human check on Preview | Includes human check on Preview |
Platform characteristics
The table below provides an eCommerce platform comparison of the deployment models.
| Characteristics | Concurrent data publishing | Cascading data publishing |
|---|---|---|
| Role of Preview environment | Preview content, configuration | Preview products, content, configuration |
| Data publishing process | From PIM Concurrently to Live, Preview always identical to Live | From PIM via Preview to Live |
| Data update process | Typical daily full update, intraday incremental updates (inserts, deletes and updates) | Typical daily full update and intraday incremental updates (updates only). Incremental update for Preview and Live are identical |
| Publishing process Preview to Live | The publishing process does not include products and typically combines content and configuration. | The publishing process always combines/bundles products, content, configuration and banners to ensure consistency |
| Typical data flow | Continuous publication of products from PIM into Preview and Live | - Monday: create a product in PIM - Tuesday: upload product from PIM to Preview - Wednesday: publish product from Preview into Production |
Supported Fredhopper business processes
The table below provides a comparison of the Business Processes as supported by Fredhopper.
| Process | Concurrent data publishing | Cascading data publishing |
|---|---|---|
| Configure Merchandising Studio | All Merchandising Studio functionality | All Merchandising Studio functionality |
| Publish Merchandising Studio configuration | Separate from the platform publishing process, this one can occur at any time. | Always bundled with the platform publishing process, including products/catalogue, content and configuration. Typical daily process |
| Create Fredhopper reports | Hourly process. Log files are fetched and processed every hour, reports are updated every hour | Hourly process. Log files are fetched and processed every hour, reports are updated every hour |
| Fully Refresh Fredhopper with data | Typical daily process. Preview and Live are fully refreshed with identical catalogues. No rollback | Typical daily process. Preview and Live are fully refreshed with different catalogues (Live = Preview yesterday, Preview = PIM yesterday). No rollback |
| Partially refresh Fredhopper with data | Partial updates include new products, product deletes and product updates. The process cannot run concurrently with a full refresh. | Partial updates include only product updates. Product inserts and deletes are not supported to guarantee consistency of products, configuration and content (e.g. banner + associated product). |
Glossary
This section explains all terms related to Fredhopper deployments:
| Term | Explanation |
|---|---|
| Production environment | All environments that have 24/7 availability |
| Live environment | Environments as experienced by your customers |
| Preview environment | Internal preview environment, used by the eCommerce team to preview site changes |
| Staging | Not used in this document as this term is often used with different meanings |
| Snapshot | A copy of a complete search or navigation index created by the indexers in regular intervals. |
| Full update | Update mode in which the existing Fredhopper index is completely removed and a new index is generated |
| Incremental updates | Update mode in which the existing Fredhopper index is preserved and only new/removed/modified items are updated |
| Fredhopper instance | One Fredhopper installation. One physical machine can host multiple instances. Multiple instances form one Fredhopper environment. |
| Load balancing | Split traffic between multiple Fredhopper Query Servers. See for details. |
Comments
0 comments
Please sign in to leave a comment.